ABSTRACT
Title of Thesis: VEGETATION AND NUTRIENT DYNAMICS
OF FORESTED RIPARIAN WETLANDS IN
AGRICULTURAL SETTINGS
Rachel Rosemarie Herbert, M.S., 2005
Thesis Directed By: Associate Professor, Dr. Andrew H. Baldwin,
Department of Biological Resources Engineering
Exported agricultural nutrients have been a major supply of excess nutrients into
the environment. Riparian wetlands may play an important part in mitigating these
nutrients and thus preventing them from migrating downstream in high concentrations.
Two riparian wetland systems, one was influenced by agriculture (agricultural site) and
one was not (reference site), were studied in Maryland. Both the plant community
structure and abiotic factors were studied. The agricultural site had lower overall species
richness and tree diversity than the reference site. Also the tree leaf litter and herbaceous
leaves at the agricultural site had higher nitrogen and phosphorus concentrations and
higher productivity based on fixed carbon than the reference site. Two nutrient
enrichment experiments were conducted at the reference site to determine the nature of
nutrient limitation. The results from these studies indicate that both plant communities
are nitrogen limited. Furthermore, individual species showed a response to increased
nutrients.
VEGETATION AND NUTRIENT DYNAMICS OF FORESTED RIPARIAN
WETLANDS IN AGRICULTURAL SETTINGS
By
Rachel Rosemarie Herbert
Thesis submitted to the Faculty of the Graduate School of the
University of Maryland, 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 C. Kangas
Dr. Gregory W. McCarty
? Copyright by
Rachel Herbert
2005
ii
Dedication
I would like to dedicate this work to my mom and dad. Without your love,
support, and encouragement I would never have made it this far. Dad, it was your
love of science and the outdoors that got me interested in environmental science in
the first place. Your knowledge and advice helped me immensely. Mom, it was your
support that kept me going all this time.  I?ll always remember you sitting in the 
living room helping me make my leaf litter traps. There are no words to express how
indebted I am to both of you. Thank you.
iii
Acknowledgements
There are many people who have been instrumental in the completion of this
project.  First, I?d like to thank my advisor, Andy Baldwin. Thanks for all of your
guidance and help.Truly you have taught me to ?think like a scientist.?  I?d also like 
to thank Greg McCarty for letting me work with you at the USDA. I appreciate all
the guidance that you?ve provided me with.  In addition I?d like to acknowledge Pat
Kangas. Your enthusiasm has always inspired me to achieve more. Thanks for all of
your input, especially in the skunk cabbage projects.
Of course I can?t forget to thank all of my cohorts in the Environmental
Quality and Wetland Ecology Laboratories. Without the help of Katie Wilson, Kate
Jacobs, Trevor Olexy, Walter Stracke, Peter Nilson, Elizabeth Urban, Jen Fetcho,
Kerry Sefton, Leticia Drakeford, Randy Roland, Krissy Rusello, Mike Egnotovich,
Alya Raphael, and Talia Chalew I would never have finished collecting and
processing all of my samples. I appreciate your dedication and willingness to go
outside with me despite grueling heat and spider infested areas!
I?d also like to thank Dr. Jonathan Angier for his hydrological input.  I would
never have been able to setup and collect data from my monitoring wells without you.
I also appreciate the countless hours you spent outside with me surveying trees.
To the rest of EQL?especially Cathleen Hapeman, Swati Mookherji, Jennifer
Klemens, Reuben Anderson, Ali Sadeghi, Emy Pfeil, and Laura McConnell?thanks
for all your encouragement throughout the years!
iv
Table of Contents
Dedication..................................................................................................................... ii
Acknowledgements......................................................................................................iii
Table of Contents......................................................................................................... iv
List of Tables ............................................................................................................... vi
List of Figures............................................................................................................. vii
Chapter 1: Introduction................................................................................................. 1
BACKGROUND............................................................................................................ 1
FORESTED RIPARIAN WETLANDS............................................................................... 2
SITE SELECTION AND LOCATION................................................................................ 3
PROJECT JUSTIFICATION AND OBJECTIVES................................................................. 8
PROJECT OVERVIEW .................................................................................................. 9
Chapter 2: Plant Community Structure and Abiotic Factors in an Agriculturally
Enriched and a Non-Enriched Wetland ...................................................................... 14
INTRODUCTION ........................................................................................................ 14
METHODS ................................................................................................................ 15
RESULTS .................................................................................................................. 18
Trees and Shrubs................................................................................................. 18
Herbaceous Vegetation ....................................................................................... 24
Abiotic Factors.................................................................................................... 24
DISCUSSION ............................................................................................................. 31
Trees and Shrubs................................................................................................. 31
Herbaceous Vegetation ....................................................................................... 33
Abiotic Factors.................................................................................................... 34
CONCLUSIONS.......................................................................................................... 35
Chapter 3: Biomass and Nutrient Concentrations of Woody Plant Litter and
Herbaceous Vegetation in Agricultural Settings......................................................... 36
INTRODUCTION ........................................................................................................ 36
METHODS ................................................................................................................ 39
RESULTS .................................................................................................................. 40
DISCUSSION ............................................................................................................. 53
CONCLUSIONS.......................................................................................................... 76
Chapter 4: The Effects of Nutrient Enrichment on the Herbaceous Vegetation of a
Forested Riparian Wetland ......................................................................................... 78
INTRODUCTION ........................................................................................................ 78
METHODS ................................................................................................................ 80
RESULTS .................................................................................................................. 83
DISCUSSION ........................................................................................................... 105
CONCLUSIONS........................................................................................................ 109
Chapter 5: The Effects of Nutrient Enrichment on Symplocarpus foetidus............. 111
INTRODUCTION ...................................................................................................... 111
METHODS .............................................................................................................. 112
RESULTS ................................................................................................................ 115
v
DISCUSSION ........................................................................................................... 126
CONCLUSIONS........................................................................................................ 130
Chapter 6: Synthesis and Conclusions..................................................................... 131
Appendix A............................................................................................................... 138
Appendix B ............................................................................................................... 142
Bibliography ............................................................................................................. 150
vi
List of Tables
Table 1-1.  Means ? SE for ammonium, nitrate, and phosphate concentrations in
groundwater in mg/L................................................................................................... 7
Table 2-1. Tree density per hectare for each species present at both sites.. .................... 22
Table 2-2. Diameter at base height (DBH) in cm for each tree species present by site.. 22
Table 2-3.  Average number of stems ? SE and height (m) ? SE for shrubs per 0.04 ha
plot at both sites. ....................................................................................................... 23
Table 2-4.  Average sapling circumference (cm) ? SE and height (m) ? SE at both sites.  .
................................................................................................................................... 23
Table 2-5. Basal area (cm2) in 0.04 ha plots and total number of trees for each species
present at both sites. ................................................................................................. 25
Table 2-6. Percent cover and frequency for herbaceous species by site. P-values are for
t-tests of site effects. ................................................................................................. 26
Table 2-7. Percent cover and frequency of abiotic factors by site................................... 29
Table 2-8. Soil concentrations of nitrate, ammonium, total nitrogen, total carbon, and the
C:N ratio at both sites. There was a significant difference between sites for total
nitrogen and total carbon concentrations????????????????..30
Table 3-1. Biomass, nutrient concentrations, nutrient loads, nutrient use efficiencies, and
element ratios for tree litter at both sites in 2002 and 2003. Significance values were
determined using t-test analyses. .............................................................................. 43
Table 3-2 Biomass, nutrient concentrations, nutrient loads, nutrient use efficiencies, and
element ratios for herbaceous vegetation at both sites in 2002 and 2003................. 54
Table 4-1. Tests of fixed effects results for all species.................................................... 85
Table 4-2. Nitrogen and phosphorus effects on percent cover and frequency of all
herbaceous species.................................................................................................... 89
Table 4-3. Tests of fixed effects results for Impatiens capensis...................................... 92
Table 4-4. Tests of fixed effects for Microstegium vimineum......................................... 93
Table 4-5. Tests of fixed effects for Polygonum punctatum............................................ 94
Table 4-6. Tests of fixed effects for Symplocarpus foetidus. .......................................... 95
Table 5-1. Tests of fixed effects for LAI, biomass, nitrogen, phosphorus, and carbon
concentrations, and N:P, C:P, and C:N ratios in Symplocarpus foetidus. .............. 117
Table 5-2. Tests of fixed effects results for length, width, and height of Symplocarpus
foetidus.................................................................................................................... 121
Table 5-3. Tests of fixed effects results for soil............................................................. 125
vii
List of Figures
Figure 1-1. Location of research sites in Laurel, Maryland............................................... 4
Figure 1-2. Close up view of the agricultural site............................................................ 10
Figure 1-3. Close up view of the reference site. .............................................................. 11
Figure 2-1. Rank-abundance curves for tree species by site............................................ 19
Figure 2-2. Comparison of frequency and cumulative percent of trees by site based on
DBH.??????????????????????????????.21 
Figure 3-1. Differences in nitrogen concentration in tree litter by species at the (a)
agricultural site and (b) reference site....................................................................... 44
Figure 3-2. Differences in phosphorus concentration in tree litter by species at the (a)
agricultural site and (b) reference site....................................................................... 45
Figure 3-3. Differences in carbon concentration in tree litter by species at the (a)
agricultural site and (b) reference site. Means with identical letters are not
significantly different at the 0.05 level by the Tukey-Kramer procedure................. 46
Figure 3-4. Differences in C:N ratios in tree litter by species at the (a) agricultural site
and (b) reference site................................................................................................. 47
Figure 3-5. Differences in C:P ratios in tree litter by species at the (a) agricultural site
and (b) reference site................................................................................................. 49
Figure 3-6. Differences in N:P ratios in tree litter by species???????????50
Figure 3-7. Change in nutrient concentrations in tree litter over time in 2002 for (a)
nitrogen, (b) phosphorus, and (c) carbon. ................................................................. 51
Figure 3-8. Change in nutrient concentrations in tree litter over time in 2003 for (a)
nitrogen, (b) phosphorus, and (c) carbon. ................................................................. 52
Figure 3-9. Differences in nitrogen concentrations in herbaceous vegetation by species in
2002 at the (a) agricultural site and (b) reference site. ............................................. 55
Figure 3-10. Differences in phosphorus concentrations in herbaceous vegetation by
species in 2002 at the (a) agricultural site and (b) reference site.............................. 56
Figure 3-11. Differences in carbon concentrations in herbaceous vegetation by species in
2002 at the (a) agricultural site and (b) reference site. ............................................. 57
Figure 3-12. Differences in N:P ratios in herbaceous vegetation by species in 2002 at the
(a) agricultural site and (b) reference site. ................................................................ 58
Figure 3-13. Differences in C:P ratios in herbaceous vegetation by species in 2002 at the
(a) agricultural site and (b) reference site. ................................................................ 59
Figure 3-14. Differences in C:P ratios in herbaceous vegetation by species in 2002 at the
(a) agricultural site and (b) reference site. ................................................................ 60
Figure 3-15. Differences in nitrogen concentrations in herbaceous vegetation by species
in 2003 at the (a) agricultural site and (b) reference site. ......................................... 61
Figure 3-16. Differences in carbon concentrations in herbaceous vegetation by species in
2003 at the (a) agricultural site and (b) reference site. ............................................. 62
Figure 3-17. Differences in C:N ratios in herbaceous vegetation by species in 2003 at the
(a) agricultural site and (b) reference site. ................................................................ 63
Figure 3-18. Differences in C:P ratios in herbaceous vegetation by species in 2003 at the
(a) agricultural site and (b) reference site. ................................................................ 64
viii
Figure 3-19. Differences in phosphorus concentrations in herbaceous vegetation by
species in 2003 at the (a) agricultural site and (b) reference site.............................. 65
Figure 3-20. Differences in N:P ratios in herbaceous vegetation by species in 2003 at the
(a) agricultural site and (b) reference site. ................................................................ 66
Figure 4-1. Nitrogen and phosphorus effects on the concentrations of (a) nitrogen, (b)
phosphorus, and (c) carbon of all herbaceous species. ............................................. 84
Figure 4-2. Nitrogen and phosphorus effects on element ratios of all herbaceous species.
................................................................................................................................... 86
Figure 4-3. Nitrogen and phosphorus effects on the nutrient loads of (a) nitrogen, (b)
phosphorus, and(c) carbon for all herbaceous species.............................................. 87
Figure 4-4. Nitrogen and phosphorus effects on the nutrient use efficiencies of (a)
nitrogen and (b) phosphorus of all herbaceous species???????????88
Figure 4-5. Nitrogen and phosphorus effects on the biomass of all herbaceous species.
................................................................................................................................... 89
Figure 4-6. Nitrogen and phosphorus treatment effects on the percent cover of (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus.............................................................................................. 91
Figure 4-7. Nitrogen and phosphorus treatment effects on the frequency of (a) Impatiens
capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus.............................................................................................. 96
Figure 4-8. Nitrogen and phosphorus treatment effects on the phosphorus concentration
in (a) Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum,
and (d) Symplocarpus foetidus. ................................................................................ 97
Figure 4-9. Nitrogen and phosphorus treatment effects on the nitrogen concentration in
(a) Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and
(d) Symplocarpus foetidus. ....................................................................................... 98
Figure 4-10. Nitrogen and phosphorus treatment effects on the carbon concentration in
(a) Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and
(d) Symplocarpus foetidus. ..................................................................................... 100
Figure 4-11. Nitrogen and phosphorus treatment effects on the N:P ratios in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus............................................................................................ 101
Figure 4-12. Nitrogen and phosphorus treatment effects on the C:N ratios in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus............................................................................................ 102
Figure 4-13. Nitrogen and phosphorus treatment effects on the C:P ratios in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus............................................................................................ 103
Figure 4-14. Nitrogen and phosphorus treatment effects on the biomass of (a) Impatiens
capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d)
Symplocarpus foetidus............................................................................................ 104
Figure 5-1. Nitrogen and phosphorus treatment effects on the leaf area indices (LAI) of
Symplocarpus foetidus. ........................................................................................... 116
Figure 5-2. Nitrogen and phosphorus treatment effects on the concentrations of (a)
nitrogen, (b) phosphorus, and (c) carbon in the leaves of Symplocarpus foetidus.
................................................................................................................................. 118
ix
Figure 5-3. Nitrogen and phosphorus treatment effects on the (a) N:P, (b) C:P, and (c)
C:N element ratios. ................................................................................................. 119
Figure 5-4. Nitrogen and phosphorus treatment effects on plant width. ....................... 120
Figure 5-5. Nitrogen and phosphorus effects on (a) plant height and (b) the length of the
youngest leaf. .......................................................................................................... 122
Figure 5-6. Nitrogen and phosphorus effects on plant biomass..................................... 123
Figure 5-7. Nitrogen and phosphorus treatment effects on (a) total nitrogen, (b)
ammonium, and (c) nitrate concentrations in soil................................................... 124
Figure 5-8. Nitrogen and phosphorus treatment effects on the carbon concentrations in
soil........................................................................................................................... 126
Figure 5-9. Nitrogen and phosphorus treatment effects on the C:N ratios in soil. ........ 127
1
Chapter 1: Introduction
BACKGROUND
The release of nutrients into natural systems is a continuing problem in the United
States (Carpenter et al. 1998). Agriculture has been a major source of this nutrient flux
(Peterjohn and Correll 1984; Brenner 1995). Today, the amount of nutrients applied as
fertilizer into agricultural fields is greater than that removed in the crop biomass
(Carpenter et al. 1998). This imbalance results in the export of excess nutrients from the
fields and may be transported to surface waters downstream via runoff and infiltration of
rain water.
Riparian wetland systems may play an important role in retaining nutrients
because of their position in the landscape (Gilliam 1994; Bischoff et al. 2001; Casey et
al. 2001). Bischoff et al. (2001) have shown that wetland soils can provide long-term
storage of nutrients while the vegetation provides short-term storage. Phosphorus
sorption and denitrification are important mechanisms of nutrient retention and removal
in the soil (Mitsch and Gosselink 1993). Wetlands can remove 90 to 100% of NO3- that
enter enters the system (Todd et al. 1984; Cooper 1990; Bischoff et al. 2001).
Researchers have also found that plant uptake is an important pathway in nutrient
cycling (Mitsch and Gosselink 1993). One study found that over 50 kg/ha/yr of N could
be removed via plant uptake in riparian forests (Todd et al. 1984). Furthermore, studies
have indicated that natural wetlands can act as sinks for phosphorus contained in
agricultural runoff (Chambers et al. 1993). Todd et al. (1984) determined that woody
vegetation in riparian forests could remove 3.8 kg/ha/yr of P while non-woody vegetation
removed 3.3 kg/ha/yr.
2
Not only does the vegetation in these systems affect the fate of these excess
nutrients, but excess nutrients can also impact the plant community structure (Verhoeven
et al. 1994). Introduction of nutrients into a natural system can limit the productivity of
some species while enhancing the productivity of others. Nutrient enrichment can
change species composition and reduce the species diversity of wetlands (Morris 1991;
Bedford et al. 1999). Nitrogen, phosphorus, or both sometimes limit the growth of
plants. Thus, the dynamics of the plant community can be altered by the import of
agricultural nutrients.
FORESTED RIPARIAN WETLANDS
Forested riparian wetlands, which may also be known as bottomland hardwood
and floodplain forests, are important because of their location in the landscape. They
occur between at the interface of terrestrial and aquatic ecosystems. These wetlands are
frequently flooded and often have high water tables because of their close proximity to
streams and rivers.
There are three major features of riparian wetlands: (1) they are linear in form
since they are close to rivers and streams; (2) they interact with energy and materials
from upstream areas more so than any other ecosystem; (3) they are linked with
upstream, downstream, upslope, and downslope areas (Mitsch and Gosselink 1993).
These characteristics mean that these areas may play an important role in treating
pollutants carried from upstream or upslope regions. Of course, their success in
attenuating or treating these pollutants is dependent on many factors, including the
existing vegetation and soils and the frequency of flooding.
The vegetation is an important factor influencing the ecosystem?s functions
3
(Silvan et al. 2004). The plants provide bank stability, flood control for areas
downstream because they slow the flow of water, temperature control via shading, leaf
litter to the system which can increase microbial activity, and wildlife habitat.
European wetlands have been largely studied and classified as either being
nitrogen or phosphorus limited (Bedford et al. 1999). However, the classification of
nutrient limitation in North American wetlands is still lacking and disputes within the
scientific community about the nature of limitation in North American wetlands are still
prevalent (Bedford et al. 1999).
SITE SELECTION AND LOCATION
I studied two riparian wetlands. One was influenced by agricultural runoff that
contained high concentrations of nutrients while the other was not directly influenced by
agriculture. Both sites are found in Laurel, Maryland, which is about 30 miles northeast
of Washington, D.C., and are part of the Chesapeake Bay Watershed. They are both
located on the grounds of Beltsville Agricultural Research Center (BARC, part of the
U.S. Department of Agriculture) and are within 1.9 km of each other (Figure 1-1). The
proximity of the two sites to each other increases the probability of having similar plant
species, soils, and hydrology at both sites. In addition, both sites were adjacent to 1st
order streams.
The agricultural site chosen for this project (known as OPE3?Optimizing
Production Inputs for Economic and Environmental Enhancment) is part of an
international research program involving many U.S. government agencies, universities,
and private industries. This site is considered to be an agricultural riparian forest buffer.
4
Figure 1-1. Location of research sites in Laurel, Maryland. Both sites are located on the grounds of
the Beltsville Agricultural Research Center (BARC, part of the USDA).
5
These buffer areas consist of trees and other vegetation and are located between
agricultural croplands and streams (or other waterbodies) or a groundwater recharge area
(Palone and Todd 1998). Such areas are very important in retaining nutrients and other
pollutants because their positions in the landscape actively slow the flow of water thereby
providing the opportunity for retention and/or biodegradation of chemicals.
The overall goal of the OPE3 project is to investigate environmental issues in both
the agricultural fields and the nearby forested riparian wetland. This site was chosen for
this project because it is surrounded by corn (Zea mays L.) fields that regularly receive
chemical fertilizer and manure for increased productivity. The area has been receiving
agricultural runoff for over 50 years. Like most forests surrounding agricultural fields,
this area is deeply fragmented in nature. The soil is mainly Typic Haplosaprist (Johnston
silt-loam series) (Angier et al. 2002) indicating that it is very poorly drained, highly
organic, and subject to flooding with a seasonally high water table at or near the surface
(Kirby et al. 1967).
This site had been instrumented with weirs, leaf litter traps, and sap flow sensors.
The height of and nutrient concentrations in groundwater had also been studied.
Moreover, surface water was collected regularly. During storm events, automated
samplers situated beside each weir were collected stream water samples while during
baseflow conditions, samples were collected manually and tested for nutrients. Thus the
hydrology has been well studied. Elevated concentrations of nitrogen in the surface
water, ground water, and the vegetation (to a limited extent) have already been
documented within this system. Preliminary vegetation samples also showed that the
leaves of Symplocarpus foetidus L. Salisb. ex Nutt. (skunk cabbage) have high
6
concentrations of phosphorus. However, the characteristics and role of vegetation in
nutrient dynamics had not been extensively studied.
The purpose of my study was to compare the plant community structures and
nutrient dynamics at OPE3 (agricultural site) to a non-agriculturally impacted wetland
(reference site). I developed several criteria for the location of the reference site. It had
to be a forested riparian wetland with similar hydrology and plant species as the
agricultural site. Skunk cabbage was used as an indicator species since it is commonly
found in areas where upwelling of water is observed. Additionally, the site could not
border any agricultural fields or other areas that may result in high nutrient loads into the
system.
The reference site, which was also located in Laurel, Maryland, was chosen
because it matched these criteria. It was a contiguous wetland forest. Upon examination
of the Prince Georges County Maryland soil survey maps (Kirby et al. 1967), the three
soil series listed at the reference site were Elkton Silt Loam, Keyport Silt Loam, and
Christiana Silt Loam. I recognize that looking at soil surveys is not a replacement for on-
site investigations of soils, but it does allow me to estimate the types of soils present.
The Elkton Silt Loam soil series consists of poorly drained silt loam over clay or clay
loam with seasonally high water tables 0-1 ft below the soil surface (Kirby et al. 1967).
The Keyport soil series contain soils that are moderately well drained, fine sandy loams,
silt loams, and silty clays that overlay thick layers of clay or silty clay that have seasonal
water table 1-2 ft below the soil surface (Kirby et al. 1967). The Christiana Silt Loam soil
series, on the other hand, is typified by well-drained soils clayey throughout the soil
profile with a thin surface layer of sandy or silt loam (Kirby et al. 1967).
7
To test the site?s suitability, preliminary groundwater and surface water samples 
were collected in June 2002 at the reference site and the agricultural site to compare
concentrations of ammonium, nitrate, and phosphate. One surface water sample was
taken and had a nitrate concentration of 0.59 mg/L. The groundwater was collected by
digging holes in the ground with a shovel until the groundwater was found. Since 2002
was a drought year, the groundwater was found at a depth between 1.5 and 2.5 feet in all
reference site holes and between 1 and 1.5 ft in depth at the agricultural site. The data
revealed that the ammonium concentrations were twice as high at the reference site as at
the agricultural site with the P-value for a t-test was slightly above 0.05 (Table 1-1).
Groundwater
AGRICULTURAL REFERENCE P
Ammonium 0.7 ? 0.10 1.4 ? 0.30 0.054
Nitrate 0.06 ? 0.03 0.05 ? 0.01 0.90
Phosphate 0.06 ? 0.01 0.04 ? 0.02 0.46
This difference may be due to differences in water saturation. The agricultural site is
generally more saturated than the reference site. This saturation would thus promote
anaerobic conditions. Since mineralization rates are usually higher in aerobic rather than
anaerobic conditions, the mineralization rates may be higher at the reference site which
could lead to higher ammonium concentrations (Richardson and Vepraskas 2001).
However, no significant difference between sites was observed for nitrate and phosphate
(Table 1-1). This insignificance may have been due to the drought. Nitrate typically is
dissolved in water and transported into nearby areas. Since the agricultural site receives
Table 1-1. Means ? SE for ammonium, nitrate, and phosphate concentrations in groundwater in
mg/L. P-values are for t-tests of site effects.
8
most of its nitrate via groundwater transport, the concentration entering the site would
have been lower due to the drought. In addition, denitrification may have been high
resulting in lower nitrate concentrations. Also, phosphorus is typically bound to soil
particles and is largely transported through the environment by surface processes, such as
runoff. During the drought there mechanism for phosphate transportation into the
wetland at the agricultural site was probably limited. Moreover, there is normally very
little phosphorus within the water column because living plants take up most of the
available phosphorus (Richardson and Vepraskas 2001).
PROJECT JUSTIFICATION AND OBJECTIVES
Excess nutrients in agricultural runoff can have detrimental effects on
downstream ecosystems. The presence of high levels of nitrogen and phosphorus in
particular is a serious problem in the United States (Carpenter et al. 1988). These
additional nutrients provide food for algae, causing algal blooms to become prevalent.
The elevated algal populations prevent light from penetrating the deeper waters. This
lack of light can then cause a decrease in submerged aquatic vegetation (SAV) since
photosynthesis rates are decreased. When these algal populations die and begin to
decompose oxygen is utilized thereby hindering the amount remaining for other
biological activities. The death of fish and other organisms, including both plants and
animals, often follow.
Therefore studies assessing the effectiveness of natural systems, like forested
riparian wetlands, to filter these excess nutrients is crucial. Likewise, it is important to
determine if the ecosystem changes in response to excess nitrogen and phosphorus to
ensure the continued removal of these nutrients. Determining the long-term impact of
9
agricultural runoff on wetland ecosystems will lead to better management practices and a
greater understanding of the role of vegetation in nutrient retention within agricultural
watersheds.
For my project, two riparian wetlands were studied. One was surrounded by
agricultural fields consisting of corn and the other was not in an agricultural setting.
Thus, the two communities could be compared in terms of plant structure and nutrient
dynamics. The objectives of my research were to:
1. determine the plant community structure of each site
2. investigate if higher nutrients at the agricultural site cause higher biomass
production and nitrogen, phosphorus, and the amount of carbon fixed in tree litter
than at the reference site
3. evaluate if the herbaceous vegetation in riparian forests in agricultural and non-
agricultural settings is limited by nitrogen, phosphorus, both, or neither (i.e., the
nature of nutrient limitation).
PROJECT OVERVIEW
Research was conducted in 2002, 2003, and 2004. In 2002, six permanently
marked 20 x 20 m plots were set up at the agricultural site (Figure 1-2). Due to time
constraints, only two plots were setup at the reference site. However, in 2003, four more
20 x 20 m plots were added at the reference site making a total of six (Figure 1-3). Two
1 x 2 m herbaceous monitoring plots were created within each of these larger plots.
Percent cover and frequency of herbaceous vegetation by species and abiotic factors, like
bare ground, coarse woody debris, etc., were determined in 2002 and 2003. Furthermore,
aboveground net primary production was estimated using the biomass of tree leaf litter
10
Figure 1-2. Close up view of the agricultural site. The yellow squares mark the locations of the 20 x
20 m plots within the wetland while the pink stars mark the locations of the surface water monitoring
stations.
11
Figure 1-3. Close up view of the reference site. They yellow squares mark the locations of the 20 x 20
m plots within the wetland while the red dots mark the locations of the monitoring wells.
and peak herbaceous aboveground biomass. Two 1 x 1 m leaf litter traps were installed
within each larger plot. Leaf litter was collected weekly during the fall season in 2002,
2003, and 2004. The litter was separated by species in 2002 but kept combined
irregardless of species in 2003. The names and authorities of all the species in this study
were obtained using the USDA Plants Database (USDA/NRCS National Plant Data
Center 2004).
Nine shallow monitoring wells were also randomly installed throughout the
reference site to measure the depth to the groundwater and to obtain samples for nutrient
analysis (Figure 1-3). Wells were not installed at the agricultural site since the
groundwater depth and nutrient concentrations had already been documented (Angier
2001; Angier et al. 2002).
12
Tree surveys were also conducted in 2002 at the agricultural site and in 2003 at
the agricultural and reference sites within each 20 x 20 m plot. Each tree, shrub, and
sapling was tagged and identified. The diameter at base height (DBH) of each tree was
measured. For shrubs, the height and stem density of each species were determined.
Sapling heights were likewise measured. Leaf area index (LAI) and photosynthetically
active radiation (PAR) were also determined.
Two fertilization experiments were conducted at the reference site. The
agricultural site was not included since plants at this location have been exposed to
agricultural inputs of nutrients for decades. The vegetation at this site therefore may not
behave in a similar manner to the addition of nutrients as non-agriculturally influenced
plants may. The objective of both experiments was to determine if the herbaceous
vegetation in riparian forests in agricultural and non-agricultural settings was limited by
N, P, both, or neither (i.e., the nature of nutrient limitation).
The first experiment occurred in 2003. Twenty 2 x 1 m plots were also randomly
set up. Five of these plots were fertilized with nitrogen; five were fertilized with
phosphorus; five received both nitrogen and phosphorus; and five received no treatment
as controls. All herbaceous vegetation within these plots was later harvested and
separated by species to determine if fertilization had any affect on biomass production,
nitrogen and phosphorus concentrations, and element ratios. All species within these
plots were harvested and examined. Aboveground net primary production, plant tissue
concentrations of nitrogen, phosphorus, and carbon, species cover, and species diversity
were measured to assess the community response at the reference site to nutrient
enhancement. The plant tissue concentrations will be measured to determine the
13
relationship between nutrient limitation and vegetation N:P ratios using the methods of
Koerselman and Meuleman (1996).
The second experiment, which occurred during the 2003-2004 growing season,
involved the fertilization of Symplocarpus foetidus only. Forty plants were randomly
chosen in a small area that was not previously utilized in another of the other studies at
the reference site. Ten plants received nitrogen, ten received phosphorus, ten received
both nitrogen and phosphorus, and ten received no treatment. The plants were harvested
so that biomass and tissue concentrations of nitrogen, phosphorus, and carbon could be
examined to determine if any treatment effects existed. Plant length, width, and height
were also measured.
14
Chapter 2: Plant Community Structure and Abiotic Factors in
an Agriculturally Enriched and a Non-Enriched Wetland
INTRODUCTION
Forested riparian wetlands are exposed to a variety of pollutants entering from
upstream and upslope. Therefore, emphasis is often placed on their ability to treat
agricultural runoff and groundwater that contains high amounts of these pollutants,
particularly nutrients. However, the influence of agriculture on the plant community
structure of these ecosystems is also important since it is a major factor that determines
the ability of a wetland to transform, store, or remove nutrients. Some studies have found
that an increase in available nutrients often causes a change in species composition,
decrease in species richness, decrease in species diversity, decrease in species evenness
(i.e. dominance of one or a few species), and invasion of exotic species (Bedford et al.
1999). Moreover, long term effects of disturbance can cause changes in species
composition, the rate of stand development and competition between plant species
(Roberts and Gilliam 1995).
Forested riparian wetlands bordering agricultural lands may be impacted by both
excess nutrients and pesticides.   These ?buffer systems? are often fragmented.  This 
characteristic may also play an important role in plant community structure. Species
richness in temperate wetlands has been related to nutrient availability, hydrology, and
topography (Bedford et al. 1999).
In this project I compared the plant community structure and abiotic factors of a
forested riparian wetland in an agriculturally influenced setting with one in a non-
15
agricultural setting. The objectives of this study were to gain a better understanding of
the influence of long term disturbance due to agricultural effluent on species richness,
plant diversity, and community evenness. I anticipated that the agricultural site would
have lower species richness, diversity, and evenness than the reference site since the
excess nutrients entering from the agricultural fields may promote the spread of a few
species that are adept at utilizing the nutrients.
METHODS
The number of trees, saplings, and shrubs were determined in each of six marked
20 x 20 m plots that were established randomly throughout the agricultural site in 2002.
However, due to time constraints, only two were placed at the reference site in 2002.
Four more plots were added in 2003 at the reference site to achieve a total of six plots.
The following criteria were used to distinguish trees, saplings, and shrubs.
Species greater than 4 m in height with a tree growth habit at maturity were considered
trees. Species less than 4 m in height with a tree growth habit at maturity were treated as
saplings. Those species with a shrub growth habit at maturity were treated as shrubs.
Within the plots, each tree was taxonomically identified according to Gleason and
Cronquist (1991) and the diameter at breast height (DBH) was measured at a height of
1.4 m above the ground. Heights were measured to the nearest 0.01 m for saplings. Each
shrub was likewise taxonomically identified and the number of stems and height to the
nearest 0.01 m were measured.
Herbaceous vegetation and abiotic factors, including bare ground, coarse woody
debris, leaf litter, and water, were surveyed in two 2 x 1 m quadrats that were
permanently marked within each of the larger plots. Most plants were identified in the
16
field based on recognition. However, if a plant could not be positively identified in the
field, a sample was taken from nearby and brought inside the laboratory and keyed out
using Gleason and Cronquist (1991). A sampling frame subdivided into 32 cells was
placed over each 2 x 1 m quadrat to determine the frequency of occurrence of individual
species and the percent coverage of each species within these cells. The percent cover is
the area of ground surface that each plant occupies if projected vertically onto the surface
within the quadrat expressed as a percentage. Percent cover values can be greater than
100% since plants often overlap each other. The cover values used were: <0.1%, 0.1%
increments from 0.1-1%, 1% increments from 1-10%, and 5% increments above 10%.
The leaf area index (LAI), which is a dimensionless expression of the foliage area
per unit area of ground surface (e.g. m2/m2), is a good indicator of aboveground biomass.
Woody LAI was measured approximately 2 m above the ground surface using a digital
canopy meter (model LAI-2000, LI-COR, Lincoln, Nebraska). Photosynthetically active
radiation (PAR), which is the light between the wavelengths of 400 and 700 nm that is
absorbed by chlorophyll, was also measured in September of 2003 at both sites using a
LI-COR 1 m quantum sensor with an LI-250 light meter (LI-COR, Lincoln, Nebraska).
The plots were divided into two equal halves (each 10 x 20 m) and measurements were
taken in the middle of each half. The wand was held approximately 2 m above the
ground while turning in a circle (radius? 1.5 m) over a 15 second averaging period.  
Nine shallow monitoring wells, which consisted of PVC pipe with slits, were
installed to a depth of about 1.1 m in topographically influenced areas (like near the base
of a hill) at the reference site. These wells allowed for the determination of the water
table height and nutrient concentrations. This depth was chosen because during the
17
drought year (2002) groundwater was obtained by digging holes in the ground and the
maximum depth to the groundwater was 0.8 m. Dr. Jonathan Angier, a USDA research
hydrologist, helped determine these key areas by looking at the landscape in the field.
No monitoring wells were placed at the agricultural site, however, because this site had
already been well instrumented with both monitoring wells and nested piezometers and
an abundance of data already existed.
The depth to the water table was determined a few times using an automated tape
that was lowered into the well and beeped when water was encountered. After the depth
had been measured, the water was pumped out of the well and the well was allowed to
replenish. It sometimes took over 24 hours for water to refill in the wells to their
previous level. Once this occurred, groundwater samples were collected and placed in
the refrigerator for 24 hours to allow suspended sediments to settle. They were then
analyzed for ammonium, nitrate/nitrite, and phosphate using a Flow Injection Analyzer
(Lachat QuikChem 8000, Milwaukee, Wisconsin).
In addition, two surface soil samples were randomly collected at a depth of 0-15
cm within each 20 x 20 m plot at both sites in 2004. These two samples were then
combined together in the field and air dried using fans in the laboratory. They were then
homogenized using a mortar and pestle. Basic soil characteristics were examined for
each site. I will use a subsample of each soil sample to determine the soil texture using a
Bouyoucos hydrometer. This method measures the density of the suspended soil, which
varies with the amount and type of particles suspended. It is imperative to recognize the
influence of organic matter on the hydrometer readings. Since it was anticipated that the
soils would contain a high amount of organic matter, the samples were oxidized prior to
18
analysis using hydrogen peroxide.
Elemental analyses were also performed on the subsamples from each soil
sample. Samples were analyzed for total organic carbon and nitrogen via dry combustion
using a LECO TruSpec 8000 CN Analyzer (St. Louis, Missouri). The Kjeldahl digestion
method was used to determine total phosphorus. Digests were then run on a Flow
Injection Analyzer (Lachat QuikChem 8000, Milwaukee, Wisconsin). Nutrients in plant
available forms were also analyzed. Exchangeable ammonium and nitrate were
determined by using a potassium chloride extraction (Mulvaney 1996). Phosphate was
determined using the dilute acid extraction procedure (also known as North Carolina or
Mehlich-1 P test) (Kuo1996).
RESULTS
Trees and Shrubs
The relative abundance and richness of tree species at both sites were examined in
2003 by plotting the abundance of each species by their rank in abundance (Figure 2-1).
The first thing to notice is that each site contains a total of seven tree species. Next, the
curves reveal that Acer rubrum L. is the most prevalent species at both sites. However,
while this species appears to be the most dominant at the agricultural site, it seems to be
co-dominating with Nyssa sylvatica Marsh and Ilex opaca Ait. at the reference site.
These graphs also reveal that there is higher species evenness at the reference site than
the agricultural site since the slope of reference site curve is less steep.
Excess nutrients in agricultural effluent may be retained by nearby riparian buffer
19
Abundance Rank
0 1 2 3 4 5 6 7 8
Pro
po
rtio
na
lA
bu
nd
an
ce
0.0
0.2
0.4
0.6
0.8
Agricultural Site
Reference Site
ACRU
ACRU
NYSY
NYSY
LIST
ILOP
UID BENI QUAL QURU
ILOP QUAL QURUQUBI
Figure 2-1. Rank-abundance curves for tree species by site. Values on the X-axis are the abundance
ranks for each species. Values on the Y-axis are the proportional abundance (number of times each
species occurred divided by the total number of trees). The reference site has greater species
evenness than the agricultural site but both communities have the same species richness.
ACRU=Acer rubrum; ILOP=Ilex opaca; LIST=Liquidambar styraciflua; NYSY=Nyssa sylvatica;
UID=Unidentified species; BENI=Betula nigra; QUAL=Quercus alba; QUBI=Quercus bicolor;
QURU=Quercus rubra.
20
vegetation, but these nutrients may negatively impact the ecosystem?s species diversity, 
which is an important measure of ecosystem health. Shannon-Wiener Diversity Indices
were also calculated for each site using the following equation:
where H ? is the value of the Shannon-Wiener diversity index, pi is the proportion of the
ith species, loge is the natural logarithm of pi, and s is the number of species in the
community. The diversity index of the trees at the agricultural site was 0.950 while the
diversity index of the reference site was 1.14. Thus, the reference site has higher tree
species diversity than the agricultural site. This higher value reflects a greater evenness
of species composition at the reference site since both sites had the same number of
species. The shrub diversity index at the agricultural site was 1.23 while that at the
reference site was 0.637.
There was also a greater frequency of smaller trees at the reference site than at the
agricultural site (Figure 2-2). The diameters of most trees at the reference site fall
between 5 and 25 cm at the reference site while those at the agricultural site fall between
30 and 45. The mode of the DBH at the reference site is 10 cm while that at the
agricultural site is 30 cm.
The density of trees on a per hectare basis was also examined for both forests
(Table 2-1). Acer rubrum was the dominant tree found at the agricultural site. There was
not a significant difference between sites or between species.
s
     H? = ? ?p i log e p i
i=1
21
Comparison of Frequency and Cumulative Percent
by Site based on DBH
0
5
10
15
20
25
30
5 10 15 20 25 30 35 40 45 50 55 60 65 70
DBH (cm)
Fre
qu
en
cy
.00%
20.00%
40.00%
60.00%
80.00%
100.00%
120.00%
Cu
mu
lat
ive
Pe
rce
nt
Agricultural Site Frequency Reference Site Frequency
Agricultural Site Cumulative % Reference Site Cumulative %
t
The amount of light penetrating through the canopy affects plant growth. PAR
measurements were taken to assess potential light differences. Recall that PAR was
taken at a height of about 2 m above the ground. Therefore this measurement would be
an assessment of light penetrating the tree canopy without the influence of existing
shrubs. No significant difference was found between sites (P>0.05) although the
agricultural site did have a higher average PAR (190.2 ? 23.27?mol s-1 m-2) than the
reference site (112.2 ? 51.79?mol s-1 m-2). LAI measurements were likewise taken. It
was found that the agricultural site did have a lower average LAI (2.6 ? 0.08) than the
reference site (3.1 ? 0.22), but was not significantly different. 
Figure 2-2. Comparison of frequency and cumulative percent of trees by site based on DBH.
Values on the X-axis are DBH classes in increments of 5 cm. Values on the left Y-axis are number
of occurrences of trees in each class while those on the right Y-axis are cumulative percent of the
observations.
22
Table 2-1. Tree density per hectare for each species present at both sites. P-values are for t-tests of
site effects. A double dash (--) indicates that the species was not present at the designated site and
therefore a t-test was not performed.
AGRICULTURAL REFERENCE
SPECIES
Mean ? SE Mean ? SE
P
Acer rubrum L. 31 ? 1.3 24 ? 1.5 0.0007
Betula nigra L. 28 -- --
Ilex opaca Ait. 5 ? 1.5 1.6 --
Liquidambar styraciflua L. -- 22 ? 4.2 --
Nyssa sylvatica Marsh 8 ? 1.4 9 ? 1.2 0.73
Quercus alba L. 22 57 --
Quercus bicolor Willd. -- 29 --
Quercus rubra L. 36 65 --
UID 25 ? 3.4 -- --
Overall Average 27 ? 1.3 20 ? 1.3 0.0004
Table 2-2. Diameter at base height (DBH) in cm for each tree species present by site. P-values are
for t-tests of site effects. A double dash (--) indicates that one or no trees of that species was present
at least one of the designated sites and therefore a t-test was not performed.
AGRICULTURAL REFERENCE
SPECIES
Mean ? SE Mean ? SE
P
Acer rubrum L. 383 ? 71.2 279 ? 53.0 0.27
Betula nigra L. 13 ? 12.5 -- --
Ilex opaca Ait. 33 ? 23.9 4 ? 4.2 0.28
Liquidambar styraciflua L. -- 83 ? 30.1 --
Nyssa sylvatica Marsh 71 ? 29.2 142 ? 73.8 0.39
Quercus alba L. 4 ? 4.2 4 ? 4.2 1.0
Quercus bicolor Willd. -- 4 ? 4.2 --
Quercus rubra L. 4 ? 4.2 4 ? 4.2 1.0
UID 17 ? 12.4 -- --
Total 525 ? 80.6 521 ? 104.2 0.98
23
However, a significant difference in DBH existed overall and for Acer rubrum
between sites (Table 2-2). It is important to note that the agricultural site has a greater
number of shrubs than the reference site (Table 2-3).
AGRICULTURAL REFERENCE
SPECIES No. of Stems Height No. of Stems Height
Ilex verticillata (L.) Gray 4 ? 1.2 3 ? 0.3 0.3 ? 0.3 4
Lyonia ligustrina (L.) DC. 1 ? 0.4 2 ? 0.5 0.2 ? 0.2 2
Unidentified Species 3 ? 1.0 4 ? 0.2 -- --
Viburnum recognitum Fern. 7 ? 3.5 3 ? 0.2 -- --
Table 2-3. Average number of stems ? SE and height (m) ? SE for shrubs per 0.04 ha plot at both
sites.
There were more saplings at the agricultural site than at the reference site (Table
2-4), but I saw many Acer rubrum seedlings (some actually growing at the base of the
Symplocarpus foetidus leaves) at the reference site which were not measured because
they we were generally less than 5 cm in height.
The basal area of the trees is another important characteristic of both systems.
While a significant difference does not exist between sites on the whole, a significant
difference does exist between sites for the most dominant tree at both sites?Acer rubrum
L. (red maple) (Table 2-5). Indeed, the average basal area of the red maples found at the
Site Species Frequency Circumference Height
Agricultural Betula nigra 3  0.04 ? 0.002 1.6 ? 0.7
Agricultural Ilex opaca 7 0.07 ? 0.02 1.0 ? 0.1
Agricultural Nyssa sylvatica 3 0.5 ? 0.2 2.9 ? 0.3
Reference Nyssa sylvatica 1 0.09 4
Table 2-4. Average sapling circumference (cm) ? SE and height (m) ? SE at both sites. There was
only one sapling at the reference site so a standard error could not be computed here.
24
agricultural site is more than double that at the reference site.
Herbaceous Vegetation
A total of 24 herbaceous species were found within the plots at the agricultural
site while a total of 35 were found at the reference site. The herbaceous plant diversity
was found to be slightly lower at the reference site (H?=1.26) than the agricultural site
(H?=1.31). Symplocarpus foetidus had the highest percent of cover and frequency at the
agricultural site (Table 2-6). While Symplocarpus foetidus contributed the highest
percent cover at the reference site, Microstegium vimineum (Trin.) A. Camus, an
introduced species that spreads very rapidly, occurred most frequently. The plant with
the second highest cover and frequency at the agricultural site was Impatiens capensis
Meerb. while that at the reference site was Microstegium vimineum. It is also important
to note that the agricultural site has significantly higher percentage of cover and
frequency of the two most dominating species at that site?Symplocarpus foetidus and
Impatiens capensis?than the reference site (Table 2-6). Both systems are dominated by
these species, but the reference site is also highly dominated by Microstegium vimineum.
Abiotic Factors
The cover and frequency of bareground, coarse woody debris, leaf litter, and
standing water were also examined (Table 2-7). It was found that bareground at the
reference site had significantly higher percentage of cover and frequency of occurrence
than the agricultural site. However, the agricultural site had higher percent cover of
coarse woody debris, leaf litter, and water.
Surface water was collected at both sites in September 2003 and analyzed for
AGRICULTURAL REFERENCE
SPECIES Num. of
trees Mean ? SE
Num. of
trees Mean ? SE
P
Acer rubrum L. 92 13398 ? 1493.4 67 6270 ? 1190.9 0.0039
Betula nigra L. 1 106 0 -- --
Ilex opaca Ait. 8 24 ? 13.9 1 0.3 ? 0.3 0.15
Liquidambar styraciflua L. 0 -- 20 2180 ? 1116.5 --
Nyssa sylvatica Marsh 17 319 ? 175.7 34 538 ? 394.5 0.62
Quercus alba L. 1 62 1 430 ? 429.7 0.43
Quercus bicolor Willd. 0 -- 1 112 ? 112.2 --
Quercus rubra L. 1 172 ? 172.3 1 557 ? 557.3 0.53
UID 4 355 ? 315.7 0 -- --
Total 124 14436 ? 1299.3 125 10087 ? 1706.6 0.070
Table 2-5. Basal area (cm2) in 0.04 ha plots and total number of trees for each species present at both sites. P-values are for t-tests of site
effects for basal area. A double dash (--) indicates that the species was not present at the designated site and therefore t-tests were not
performed.
COVER (%) FREQUENCY (%)SPECIES
AG. REF. P AG. REF. P
Apios americana Medik. 0 0.1 ? 0.09 0.19 0 2 ? 1.6 0.27
Arisaema triphyllum (L.) Schott 0 4 ? 1.9 0.060 0 11 ? 4.4 0.033
Boehmeria cylindrica (L.) Sw. 0.4 1 ? 0.8 0.57 3 ? 3.1 2 ? 1.6 0.77
Carex folliculata L. 0 0.2 0.34 0 1 ? 0.5 0.34
Carex intumescens Rudge 0.8 ? 0.54 0.5 ? 0.42 0.68 2 ? 1.6 2 ? 1.6 1.0
Carex lurida Wahlenb. 0 0.4 0.34 0 2 ? 1.6 0.34
Carex sp. 1 0 0.4 0.34 0 1 ? 1.0 0.34
Carex sp. 2 0 0.4 0.34 0 1 ? 1.0 0.34
Carex sp. 3 0 0.3 0.34 0 1 ? 0.8 0.34
Cinna arundinacea L. 2.2 ? 2.08 5.3 ? 3.9 036 13.3 ? 2.3 25 ? 10.9 0.34
Claytonia virginica L. 0.9 ? 0.3 2.7 ? 2.1 0.40 8.9 ? 2.4 12.5 ? 6.4 0.60
Cyperaceae 0.3 ? 0.3 0 0.22 1 ? 1.1 0 0.24
Dioscorea villosa L. 0 0.7 ? 0.50 0.14 0 2 ? 1.3 0.11
Duchesnea indica (Andr.) Focke 0.4 0 0.34 4 ? 3.4 0 0.34
Galium tinctorium L. 0.7 ? 0.43 0 0.15 4 ? 2.1 0 0.087
Table 2-6. Percent cover and frequency for herbaceous species by site. P-values are for t-tests of site effects.
(Page 1 of 3)
27
COVER (%) FREQUENCY (%)SPECIES
AG. REF. P AG. REF. P
Grass1 0 1 ? 0.47 0.047 0 10 ? 4.4 0.045
Impatiens capensis Meerb. 6 ? 1.5 2 ? 0.87 0.015 39 ? 9.8 9 ? 3.4 0.014
Juncus effusus L. 0 0.7 ? 0.47 0.18 0 2 ? 1.1 0.13
Leersia oryzoides (L.) Sw. 0.2 ? 0.1 0.1 0.54 1 ? 0.6 0.3 ? 0.3 0.24
Microstegium vimineum (Trin.) A. Capus 0 17 ? 4.2 0.0020 0 61 ? 10.1 <0.0001
Moss 2 ? 0.7 0.8 ? 0.56 0.17 13 ? 4.5 3 ? 2.1 0.059
Onoclea sensibilis L. 0 1 0.34 0 0.3 ? 0.3 0.34
Osmunda cinnamomea L. 4 ? 2.3 10 ? 6.4 0.41 15 ? 8.1 15 ? 8.3 0.97
Panicum clandestinum (L.) 0 0.4 ? 0.26 0.14 0 2 ? 1.1 0.17
Parthenocissus quinquefolia (L.) 0.1 0.1 1.0 1 ? 1.0 0.3 ? 0.3 0.48
Peltandra virginica (L.) Schott 0.4 ? 0.41 0 0.31 1 ? 0.6 0 0.19
Pilea pumilia (L.) Gray 0.2 ? 0.11 0.3 ? 0.26 0.66 2 ? 1.1 1 ? 0.8 0.58
Polygonum arifolium L. 4 ? 3.3 2 ? 1.7 0.64 9 ? 7.5 3 ? 1.7 0.47
Polygonum punctatum Ell. 1 ? 0.5 0.7 ? 0.43 0.63 15 ? 6.4 6 ? 3.3 0.25
Polygonum sagittatum L. 0 0.1 0.34 0 0.3 ? 0.3 0.34
Table 2-6 (Page 2 of 3)
(Continued)
28
COVER (%) FREQUENCY (%)SPECIES
AG. REF. P AG. REF. P
Rubus L. 0.7 ? 0.47 0 0.18 3 ? 2.0 0 0.18
Sagittaria latifolia Willd. 0 0.04 0.34 0 1 ? 0.5 0.34
Smilax rotundifolia L. 2 ? 0.6 0.4 ? 0.42 0.17 8 ? 2.9 2 ? 1.3 0.058
Spp. 1 0 0.1 0.34 0 1 ? 1.0 0.34
Spp. 2 0.1 ? 0.09 0 0.19 6 ? 5.2 0 0.26
Spp. 3 0 0.1 0.34 0 0.3 ? 0.3 0.34
Spp. 4 0.5 ? 0.29 0 0.14 6 ? 3.9 0 0.14
Symplocarpus foetidus (L.) Salisb. Ex Nutt. 71 ? 6.7 46 ? 8.5 0.031 94 ? 6.0 58 ? 11.6 0.015
Thalictrum pubescens Pursh 0 0.3 ? 0.26 0.22 0 2 ? 1.3 0.26
Toxicodendron radicans (L.) Kuntze 0.3 0.3 ? 0.13 1.0 3 ? 2.6 2 ? 1.1 0.79
Unidentified Species 2.1 ? 0.48 0 0.061 59 ? 8.8 0 <0.0001
Violet 0 0.3 0.34 0 1 ? 0.5 0.34
Table 2-6 (Page 2 of 3)
(Continued)
Table 2-7. Percent cover and frequency of abiotic factors by site. P-values are for t-tests of site effects. Bareground is ground
devoid of vegetation; coarse woody debris includes fallen trees and branches; leaf litter is litter cover the ground; water
includes standing and flowing water.
COVER (%) FREQUENCY (%)
AG. REF. P AG. REF. P
Bareground 0.8 ? 0.56 19 ? 4.7 0.0022 4 ? 2.5 25 ? 4.4 0.0003
Coarse Woody Debris 4 ? 1.2  2 0.4955 22 ? 5.7 2 ? 2.1 0.0061
Leaf Litter 10 ? 2.7 7 ? 4.2 0.3949 63 ? 12.3 14 ? 8.4 0.0030
Water 8 ? 5.8 2 ? 1.4 0.3119 9 ? 6.5 2 ? 1.4 0.28
30
nitrate. Although no significant difference was detected between sites (P=0.13), the
agricultural site (1.3 ? 0.30) had almost twice the amount of nitrate as the reference site
(0.7 ? 0.03).
Results from the monitoring wells revealed that in July of 2003, the groundwater
was found at an average depth of 18 cm while in September it was at a depth of 38 cm.
These shallow depths (compared to 2002 when I had to dig more than 1 m in depth to
reach the groundwater) were expected since the area is a wetland and Maryland
experienced its wettest year on record in 2003 (NOAA/USHCN). These water levels are
also within the range for the soil series present there (i.e. soil with water tables between
0-5 ft) (Kirby et al. 1967).
A loamy soil texture was present at the agricultural site while a silt loam soil
texture was present at the reference site. Nutrient analyses revealed that both nitrogen
and carbon concentrations were significantly higher in the agricultural soil than the
reference soil (Table 2-8). However, no differences in nitrate and ammonium
concentrations or the C:N ratio were detected (Table 2-8). The C:N ratios were both
about 15.6.
Parameter AgriculturalSite ReferenceSite P
Nitrate (mg/L) 1.83 ? 0.5 1.42 ? 0.6 0.59
Ammonium (mg/L) 6.23 ? 1.4 4.21 ? 0.6 0.20
N (%) 0.57 ? 0.6 0.31 ? 0.1 0.050
C (%) 8.97 ? 1.7 4.81 ? 0.96 0.060
C:N 15.59 ? 0.6 15.69 ? 0.7 0.92
Table 2-8. Soil concentrations of nitrate, ammonium, nitrate, and carbon and the C:N
ratio at both sites. There was a significant difference between sites in total nitrogen
and total carbon concentrations.
31
DISCUSSION
Trees and Shrubs
The prevalence of certain tree species can often predict the nature of flooding in
an area (Kozlowski 2002). Both systems are largely dominated by Acer rubrum.
Kozlowski found that Acer rubrum usually dominates riparian forests with short periods
of flooding (2002). Nyssa sylvatica was also prevalent at both sites along with Quercus
alba. These species often occur on high ridges that have very short flooding regimes
(Kozlowski 2002). Lyquidambar styraciflua likewise occurs at the reference site. This
species usually occurs on ridges that experience short flooding (Kozlowski 2002).
The DBH of trees found at the agricultural site was significantly greater than at
the reference site. This site effect is due to the significantly higher DBH of Acer rubrum
L. (red maple) at the agricultural site. Most trees at the agricultural site had DBH values
of 30 cm while most at the reference site had values of 10 cm. The nutrients entering the
agricultural system may be causing the trees to grow at a faster rate than those at the
reference site. The productivity of this system may indeed be increasing due to these
elevated nutrients. These results have also been found by Tessier and Raynal (2003).
They found that Acer rubrum trees may be limited by nitrogen and therefore may
experience higher growth rates in environments where nitrogen is prevalent.
It was somewhat surprising that the sites were not significantly different in terms
of tree density. It was anticipated that the reference site, which is a large expanse of
contiguous forest, would have a greater density than the agricultural site, which is deeply
fragmented in nature.
Both light and nutrient availability influence plant growth and survival. To
32
determine the influence of light on each ecosystem, PAR and LAI measurements were
collected. Neither PAR nor LAI were significantly different between sites. Recall that
the agricultural site was found to have lower tree diversity but higher shrub diversity. I
took both measurements at a height that included the shrub canopy. The higher number
of shrubs at the agricultural site may be influencing both of these measurements causing
an insignificant difference between sites. Thus, it may be inferred that most differences
between sites are due more to discrepancies in nutrient regimes than light regimes.
However, since Hurricane Isabel had impacted the area just a few days prior to my
measurement dates, some leaves and trees had fallen. In fact, two trees at each site had
fallen near plots.  My PAR values (112.2 ? 51.79?mol s-1 m-2 for the agricultural site and
190.2 ? 23.27?mol s-1 m-2 for the reference site) are consistent with other forest-field
systems. Cadenasso et al. (1997) studied PAR in two well-developed upland forests and
compared measurements along the forest-field edge boundaries. They found PAR values
ranging between 50 and 150 ?mol s-1 m-2 within the forest (Cadenasso et al. 1997).
No difference in the density of the systems was found. It is beneficial to have
similar values between sites to reduce the error that this may cause in other studied
parameters like herbaceous plant growth. However, these results were not anticipated.
As stated previously, the agricultural site is fragmented and has a high potential for
disturbance. Wind can penetrate this ecosystem more easily than at the reference site and
could potentially cause more trees to fall. Therefore it was thought that the trees at the
agricultural site would have a higher turnover rate resulting in less density. However, it
is important to note that the agricultural site has a greater number of shrubs than the
reference site. These shrubs may be blocking light to the lower canopy, which includes
33
seedlings. This may be one reason for the lower regeneration of trees at the agricultural
site. Furthermore, since the reference site has a larger population of smaller trees, the
chances of these trees surviving environmental and climatic changes to reach maturity is
also greater than at the agricultural site.
Furthermore, recall that there was a significantly greater amount of leaf litter
present on the ground of the agricultural site than the reference site. This litter can
further block sunlight from seedlings. Also, it has been found that litter can actually
compete with plants for nutrients (Xiong, Nilsson, and Johansson 2001). The litter may
be one reason for the lower regeneration of trees at the agricultural site. Furthermore
since the reference site has a larger population of smaller trees, the chances of these trees
surviving environmental and climatic changes to reach maturity is also greater than at the
agricultural site.
Although the overall basal area of the trees was not significantly different
between sites, it was significantly different for Acer rubrum. This result further confirms
that the Acer rubrum trees at the agricultural site are larger than those at the reference
site. Tessier and Raynal (2003) studied the N:P ratios in plant foliage. They concluded
that Acer rubrum was nitrogen limited and in the presence of unlimited quantities of
nitrogen, Acer rubrum trees would experience higher growth rates (Tessier and Raynal
2003). Since nitrogen is readily available at the agricultural site, these trees are probably
growing at a faster rate than those found at the reference site.
Herbaceous Vegetation
The herbaceous vegetation at the reference site had a slightly lower diversity
index than that at the agricultural site. This result was the opposite of what I had
34
anticipated. It is probably caused by a larger number of minor herbaceous species at the
reference site. The reference site actually had higher herbaceous species richness than
the agricultural site. This result was anticipated since the nutrients entering the
agricultural site from surrounding fields probably have promoted the growth of species
capable of adjusting to and thriving in communities with high nutrient levels. Indeed,
nutrient enrichment in several wetland types has resulted in a decrease in species richness
(Vermeer and Berendse 1983; Bedford et al. 1999).
Both systems are dominated by a few herbaceous species. Microstegium
vimineum occurred most frequently at the reference site. This species has many qualities
that make it invasive and able to adapt to disturbance quite easily. These characteristics
include easily adapting to changing environmental conditions, an annual life history, the
production of large quantities of seeds, a lack of predators or pathogens, and clonal
growth to allow for the of necessary resources.
Abiotic Factors
The higher concentrations of nitrate found in the surface water at the agricultural
site were expected since these levels are often increased in waterways bordering
agricultural fields (Todd et al. 1984; Stone et al. 1998; Angier et al. 2002). I also
expected the total nitrogen and carbon in the soils to be higher at the agricultural site than
at the reference site because I thought productivity and nutrient concentrations in the
plant litter would be higher at the agricultural site resulting in more litter with elevated
concentrations being returned to the ecosystem.
The higher concentrations of nitrogen in the soil at the agricultural site may be
due to the elevated concentrations of nitrogen entering from the agricultural fields. The
35
higher carbon concentrations may be related to the higher nitrogen concentrations.
Perhaps the additional nitrogen stimulates soil microorganisms involved in the
breakdown of organic matter. Carbon would therefore be released. Fenn et al. (1998)
found that soils with C:N ratios greater than 20 were more effective in retaining nitrogen
than those less than or equal to 10. This indicates that the soils at both sites may
potentially be effective at retaining nitrogen.
CONCLUSIONS
Although the two ecosystems are similar are many respects, there are some key
differences. The agricultural site has lower species richness and is less diverse than the
reference site. Although both systems are dominated by Acer rubrum, the reference site
appears to be dominated with two additional species. The increased nutrients entering the
agricultural system could be causing these results by selecting for species that are better
able to cope with disturbance. Acer rubrum may be one of these species since its
diameter and basal area are so significantly higher at the agricultural site than at the
reference site. It is also interesting that Microstegium vimineum, an introduced species, is
so prevalent at the reference site.
36
Chapter 3: Biomass and Nutrient Concentrations of Woody
Plant Litter and Herbaceous Vegetation in Agricultural
Settings
INTRODUCTION
The nutrient enrichment of natural waters has been due in part to agricultural
practices (Peterjohn and Correll 1984; Schr?der et al. 2004).  Pollution due to elevated
nitrogen is of particular concern since it can cause eutrophication problems downstream.
While best management practices (BMPs), have been developed to reduce the export of
agricultural effluent, it is unrealistic to presume that nutrients are not entering adjacent
areas. One BMP that has received much attention is the use of buffer strips. These
vegetated areas can be composed of herbaceous vegetation and/or woody vegetation
(trees and shrubs). The vegetation in these areas acts to slow the flow of water thereby
trapping suspended solids and allowing for plant uptake of nutrients to occur. The soils
in these areas are also important. Bischoff et al. (2001) found that wetlands stored 99%
of their nitrogen in peat, organic soil, and mineral soil and <0.1% in foliage and
herbaceous vegetation and woody tissues.
While it is important to determine the effectiveness of these buffer systems to
mitigate nutrients, it is equally imperative to study the impacts of the nutrients on these
ecosystems. Nutrient cycling is an important function of ecosystems that may be affected
by elevated nutrients from agriculture.
Plant uptake is an important pathway in the cycling of nutrients (Mitsch and
Gosselink 1993). The concentrations of nitrogen, phosphorus, and carbon in herbaceous
37
vegetation and tree leaf litter are important since they are ultimately returned to the
ecosystem at the end of the growing season, incorporated into the soil via decomposition,
and made available for use by soil microbes.
The concentrations of nutrients in herbaceous biomass and tree litter are an
important factor in nutrient cycling. The resorption of nutrients is an important
conservation mechanism. It reflects the degree to which nutrients are retained by the
plant. The concentrations of nutrients resorbed by vegetation influences competition,
nutrient uptake, and productivity (Killingbeck 1996).
Calculating nutrient use efficiencies (NUE), which are the litter biomass per unit
mass of nutrient, provide an insight into the role of nutrients in ecosystems. There has
been some disagreement within the scientific community about whether areas of high
nutrient availability result in lower or higher nutrient use efficiencies (Killingbeck 1996).
Several studies have found that these efficiencies are lower in nutrient rich areas
(Vitousek 1982; Boerner 1984; Anderson and Eickmeier 2000; Koutroubas et al. 2000).
Chapin and Kedrowski (1983) found that four Alaskan tree species growing in low
nutrient environments did not resorb a greater amount of nitrogen and phosphorus than
those in high nutrient areas.
Determining how the import of agricultural nutrients impact the vegetation
component of nutrient cycling in buffer systems is important since plants play an
essential role in pollution mitigation. In my research, I studied the vegetation component
of nutrient cycling in two forested riparian wetlands: one wetland was influenced by
agriculture (the agricultural site) while the other was not (the reference site).
38
I developed two objectives for this project. The first was to determine if nutrients
entering from surrounding agricultural fields were impacting the concentrations of
nitrogen, phosphorus, and carbon in tree litter and herbaceous vegetation collected at the
agricultural site by comparing their concentrations to those found at the reference site.
Once these concentrations were determined, nutrient loads, NUEs, and element ratios
(N:P, C:N, and C:P) were calculated and statistical tests were performed to determine if
any differences existed between sites and species. The second objective was to determine
if nitrogen, phosphorus, and carbon concentrations in the litter changed during the period
of senescence. If the nutrient concentrations at the beginning of the season were more
elevated at the agricultural site than the reference site I would presume that the nutrient
concentrations in green tree leaves were also higher. A change in the nutrient
concentrations during the period of senescence would imply that the trees are effectively
resorbing the nutrients from their leaves.
Four main hypotheses were developed to address these objectives. First, higher
nutrient availability at the agricultural site would result in elevated nutrient
concentrations and higher biomass production in both leaf litter and herbaceous
vegetation. Second, the trees found at the reference site would have higher NUEs
because the nutrients are presumably more limited in this environment and these plants
would probably need to conserve nutrients more than those in the agricultural site. Third,
the nutrient concentrations would differ by species within the woody and herbaceous
communities since some species may be more adept at resorbing nutrients than others.
Fourth, I anticipated that the nutrient concentrations would change over time since the
leaves that abscise late in the season are attached to the trees for a longer period of time
39
thereby providing a greater opportunity for nutrient retention.
METHODS
Field and Laboratory Techniques. In 2002, six 20 x 20 m (400 m2) plots were randomly
established at the agricultural site and two at the reference site. In 2003 four more plots
were added at the reference site. Two 1 m2 leaf litter traps were randomly placed within
each of the larger plots. The traps were constructed from PVC pipe and nylon screening
and positioned approximately 0.5 m above the ground. In 2002 litter was collected four
times (November 6, 15, 20 and December 23) and was kept separated by date. Litter was
likewise collected weekly in 2003 from October 10 through November 18 and on January
13 in 2004. The litter was dried at 60?C until constant weight was achieved. In 2002 the
litter was separated by species, but combined in 2003 because I did not expect
intraspecific variation to change between years. The litter from both years was weighed
and homogenized using a Wiley Mill.
A 2.5 m2 ring was randomly placed at seven locations within each 400 m2 plot so
that herbaceous vegetation could be harvested in the middle of July in 2002 and at the
beginning of August in 2003. All herbaceous vegetation was separated by species in situ
at the time of collection, dried at 60?C until constant weight was achieved, weighed, and
homogenized using a Wiley Mill.
Total nitrogen and carbon concentrations in the litter and herbaceous vegetation
were determined via dry combustion using a LECO Tru-Spec CN Analyzer (St. Louis,
Missouri). Total phosphorus concentrations were determined using a modified (instead
of adding 3.5 mL of sulfuric acid to each sample, I added 5 mL) Kjeldahl digestion
method. Digests were then run on a Flow Injection Analyzer (Lachat QuikChem 8000,
40
Milwaukee, Wisconsin). In 2003 only the litter obtained between October 10 and
November 14 was analyzed for nutrient concentrations since litter collected later in the
season did not have enough biomass for analysis.
It has been suggested that looking simply at nutrient concentrations in litter may
lead to biased results since specific leaf mass (SLM) changes during senescence (Aerts
1996). Nonetheless changes in SLM are generally small (Chapin and Moilanen 1991), so
I assumed that comparisons based on concentration were reasonable. However, I also
converted nutrient concentrations for each sample into units of mass by multiplying the
concentration by the sample biomass and I then used these values to calculate nutrient
loads, NUEs, and element ratios.
Statistical Analyses. The statistical program SAS (v. 8 for Windows, SAS Institute,
Cary, NC, USA) was used in all statistical analyses. Comparisons of biomass, nutrient
concentrations, nutrient loads, NUEs, and element ratios were compared between sites for
both years using t-tests. During 2002 differences between species both within and
between sites were also examined using tests of fixed effects. Changes in nutrient
concentrations during the period of senescence were also examined separately for 2002
and 2003 by performing repeated measures analyses. I did not average these two years
together because in 2002 I had insufficient plot replication at the reference site.
Moreover, 2002 was a drought year while 2003 was a ?normal? year.Comparisons were
also made for each year within sites between dates using repeated measure analyses and
between sites on specific dates using the Tukey-Kramer procedure.
RESULTS
Results varied depending on the year examined. In 2003, the nitrogen
41
concentration in tree leaf litter was significantly higher at the agricultural site than at the
reference site (Table 3-1). The concentration of phosphorus was not significantly
different between sites in 2002; however, it was marginally significantly higher at the
agricultural site for 2003 (Table 3-1). Unlike nitrogen and phosphorus, the concentration
of carbon was statistically higher at the agricultural site than at the reference site during
both years (Table 3-1).
In 2003, the nutrient use efficiency of nitrogen, phosphorus, and carbon were all
significantly higher in the tree litter from the reference site than at the agricultural site in
2003 (Table 3-1). Although neither the N:P nor the C:P ratios was significantly different
between sites in 2003 (Table 3-1). The C:N ratio was significantly higher at the reference
site than at the agricultural site (Table 3-1).
In 2003, the leaf litter biomass was not significantly different between sites at the
0.05 significance level using t-tests (Table 3-1). However, it was slightly higher at the
reference site than at the agricultural site during both years. Although significant
differences were observed in nutrient concentrations, none were detected for the nutrient
loads of nitrogen, phosphorus, and carbon in 2003 (Table 3-1).
However, in 2002, there were no observed differences between sites for tree litter
biomass, nitrogen and carbon concentrations, nitrogen and phosphorus NUEs, nitrogen,
phosphorus, and carbon loads, and N:P, C:P, and C:N ratios (Table 3-1). In fact, only
two significant differences were found. The concentration of carbon was significantly
higher at the agricultural site than at the reference site (Table 3-1).
Significant differences within sites between species existed for nitrogen,
phosphorus, and carbon concentrations as well as C:N ratios in tree leaf litter (Figure 3-1;
42
Figure 3-2; Figure 3-3; Figure 3-4). C:P ratios differed by species at the agricultural site
but not at the reference site (Figure 3-5). A significant species effect and an insignificant
site effect were found for N:P ratios so the species effect was averaged across sites
(Figure 3-6). Only a few differences in the studied parameters were found when
comparing species between sites. Both of these differences involved carbon. The
concentration of carbon in Liquidambar styraciflua and Quercus rubra were significantly
higher at the agricultural site than at the reference site (Figure 3-3). Likewise, the C:N
ratio in Quercus falcata was significantly higher at the agricultural site than at the
reference site (Figure 3-4).
Differences between species both within and between sites were examined in
2002 only. Statistical tests of fixed effects revealed that a significant site*species
interaction existed for nitrogen so interaction effects are presented.
Changes in nutrient concentrations during the senescence (i.e. fall) season were
examined for both years. In 2002, no changes in nitrogen (Figure 3-7a), phosphorus
(Figure 3-7b), or carbon (Figure 3-7c) concentrations over time at either site and between
sites by date were detected. In 2003, the concentration of nitrogen at the agricultural site
did change significantly over time whereas it did not at the reference site (Figure 3-8a).
From October 10 through November 5 in 2003 a decreasing trend was observed at the
agricultural site (Figure 3-8a). However, surprisingly on the last date a significant
increase occurred. Significant differences were also observed between sites in 2003
throughout the entire season except for on October 29 and November 5.
Agricultural Site Reference Site Agricultural Site Reference SiteParameter
2002 2002 P 2003 2003 P
Biomass
(kg/ha) 3282.63 ? 383.6 4645.58 ? 1192.7 0.1799 1981.75 ? 216.9 2450.1 ? 197.9 0.14
N (%) 0.80 ? 0.03 0.74 ? 0.04 0.3707 1.16 ? 0.03 0.99 ? 0.03 0.0023
P (%) 0.05 ? 0.01 0.06 ? 0.004 0.7761 0.08 ? 0.005 0.07 ? 0.002 0.055
C (%) 51.50 ? 0.1 50.35 ? 0.1 0.0009 51.03 ? 0.4 49.88 ? 0.2 0.025
N (kg/ha) 26.17 ? 3.3 34.84 ? 10.7 0.3100 22.54 ? 2.2 24.4 ? 2.3 0.57
P (kg/ha) 1.71 ? 0.1 2.77 ? 0.9 0.4461 1.58 ? 0.2 1.68 ? 0.1 0.70
C (kg/ha) 1689.68 ? 195.4 2338.42 ? 604.6 0.2052 1010.61 ? 108.2 1222.1 ? 98.5 0.18
N?NUE 127.60 ? 4.7 135.78 ? 6.9 0.4059 87.28 ? 2.5 101.2 ? 2.9 0.0048
P?NUE    1908.21 ? 148.9 1767.19 ? 134.5 0.6315   1270.92 ? 75.5 1468.1 ? 46.2 0.050
N:P Ratio 15.28 ? 1.6 12.99 ? 0.4 0.4490 14.68 ? 1.1 14.53 ? 0.4 0.90
C:N Ratio 65.72 ? 2.4 68.36 ? 3.3 0.5890 44.5 ? 1.1 50.47 ? 1.4 0.0082
C:P Ratio 982.80 ? 76.5 890.17 ? 66.0 0.5412 647.36 ? 36.8 732.46 ? 22.5 0.077
Table 3-2. Biomass, nutrient concentrations, nutrient loads, nutrient use efficiencies, and element ratios for tree litter at both sites in 2002 and 2003.
Significance values were determined using t-test analyses.
Ni
tro
ge
n(
%)
0.0
0.5
1.0
1.5
2.0
2.5
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
Ni
tro
ge
n(
%)
0.0
0.5
1.0
1.5
2.0
2.5
b b ab
ab
b b
a
ab b
b
ab
ab ab a ab ab ab a
(a)
(b)
Figure 3-1. Differences in nitrogen concentration in tree litter by species at the (a) agricultural site and (b) reference site. Means with identical letters
are not significantly different at the 0.05 level by the Tukey-Kramer procedure. There were no significant differences between sites at the 0.05 level by
the Tukey-Kramer procedure within species. The horizontal lines represent resorption levels for deciduous tree species proposed by Killingbeck (1996).
ACRU=Acer rubrum; LIST=Liquidambar styraciflua; LITU=Liriodendron tulipifera; NYSY=Nyssa sylvatica; PRSE=Prunus serotina;
QUAL=Quercus alba; QUBI=Quercus bicolor; QUFA=Quercus falcata; QULY=Quercus lyrata; QUPH=Quercus phellos; QURU=Quercus rubra.
Ph
osp
ho
rus
(%
)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
Ph
osp
ho
rus
(%
)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
(a)
(b)
b ab
ab
ab ab b b b
a
b b ab
ab
ab
a ab
ab
ab
Figure 3-2. Differences in phosphorus concentration in tree litter by species at the (a) agricultural site and (b) reference site. Means with identical
letters are not significantly different at the 0.05 level by the Tukey-Kramer procedure. There were no significant differences between sites at the 0.05
level by the Tukey-Kramer procedure within species. The horizontal lines represent resorption levels for deciduous tree species proposed by
Killingbeck (1996). Values less than 0.05 % represent complete resorption while those above 0.08 % represent incomplete resorption. Values between
these concentrations represent intermediate resorption. ACRU=Acer rubrum; LIST=Liquidambar styraciflua; LITU=Liriodendron tulipifera;
NYSY=Nyssa sylvatica; PRSE=Prunus serotina; QUAL=Quercus alba; QUBI=Quercus bicolor; QUFA=Quercus falcata; QULY=Quercus lyrata;
QUPH=Quercus phellos; QURU=Quercus ruba.
Ca
rbo
n(
%)
40
50
60
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
Ca
rbo
n(
%)
40
50
60
ac
b
ab
bc ab
ab ab ab a
ab* abc*
bc c
a b a a ab*
(a)
(b)
Figure 3-3. Differences in carbon concentration in tree litter by species at the (a) agricultural site and (b) reference site. Means with identical letters
are not significantly different at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them indicate a significant difference
exists between sites at the 0.05 level by the Tukey-Kramer procedure within species. ACRU=Acer rubrum; LIST=Liquidambar styraciflua;
LITU=Liriodendron tulipifera; NYSY=Nyssa sylvatica; PRSE=Prunus serotina; QUAL=Quercus alba; QUBI=Quercus bicolor; QUFA=Quercus
falcata; QULY=Quercus lyrata; QUPH=Quercus phellos; QURU=Quercus rubra.
C:
N
0
20
40
60
80
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
C:
N
0
20
40
60
80
(a)
(b)
a
abc
abc
bc bc
abc
ab*
bc
c
ab a ab ab ab
ab
b
ab ab
Figure 3-4. Differences in C:N ratios in tree litter by species at the (a) agricultural site and (b) reference site. Means with identical letters are not
significantly different at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them indicate a significant difference exists
between sites within species at the 0.05 level by the Tukey-Kramer procedure. ACRU= Acer rubrum; LIST=Liquidambar styraciflua;
LITU=Liriodendron tulipifera; NYSY=Nyssa sylvatica; PRSE=Prunus serotina; QUAL=Quercus alba; QUBI=Quercus bicolor; QUFA=Quercus
falcata; QULY=Quercus lyrata; QUPH=Quercus phellos; QURU=Quercus rubra.
48
The concentration of phosphorus on the first sampling date in 2003 at the
agricultural site was significantly greater than that on subsequent sampling dates (Figure
3-8b). At the agricultural site, a decreasing trend was observed for this year in general
except for the last sampling date. There were also significant differences between sites at
the beginning and end of the season. In addition, the carbon concentrations in 2003
showed that a significant increase occurred at the beginning of the season at the
agricultural site (Figure 3-8c). Nevertheless, no difference was observed at the reference
site throughout the season (Figure 3-8c). There was a difference between sites in the
middle of the season at the agricultural site but no difference at the reference site.
Overall comparisons by site for herbaceous vegetation during both years revealed
little difference between the two sites. Indeed, in 2002, the only significant difference
that existed was in the N:P ratio, which was found to be significantly higher at the
reference site than at the agricultural site (Table 3-2). In 2003, the concentration of
carbon was marginally significantly higher at the agricultural site than at the reference
site (Table 3-2).
In 2002, differences existed at both sites between herbaceous species. The nitrogen
concentrations differed by species within both the agricultural and reference sites (Figure
3-9). The phosphorus concentrations differed by species within the agricultural site but
not within the reference site (Figure 3-10). Moreover, the phosphorus concentrations
were significantly higher in Impatiens capensis Meerb., Polygonum arifoliam L, and
Symplocarpus foetidus (L.) Salisb. ex Nutt at the agricultural site than at the reference site
(Figure 3-10). The carbon concentrations also differed between species within the
agricultural site but not within the reference site (Figure 3-11). In addition, differences in
C:
P
0
200
400
600
800
1000
1200
1400
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
C:
P
0
200
400
600
800
1000
1200
1400
(a)
(b)
a
ab
ab
ab ab
ab
a
ab
b
a a a
a
a
a
a
a a
Figure 3-5. Differences in C:P ratios in tree litter by species at the (a) agricultural site and (b) reference site. Means with identical letters are not
significantly different at the 0.05 level by the Tukey-Kramer procedure. There were no significant differences between sites within species at the 0.05
level by the Tukey-Kramer procedure. ACRU= Acer rubrum; LIST=Liquidambar styraciflua; LITU=Liriodendron tulipifera; NYSY=Nyssa sylvatica;
PRSE=Prunus serotina; QUAL=Quercus alba; QUBI=Quercus bicolor; QUFA=Quercus falcata; QULY=Quercus lyrata; QUPH=Quercus phellos;
QURU=Quercus rubra.
Species
ACRU LIST LITU NYSY PRSE QUAL QUBI QUFA QULY QUPH QURU
N:
P
0
5
10
15
20
25
ab ab ab
ab ab
ab
a
ab b
ab
ab
Figure 3-6. Differences in N:P ratios in tree litter by species. Since the site*species interaction was insignificant (P=0.6147),the species effect is shown
averaged across site. A significant species effect (P=0.0412) and an insignificant site effect (P=0.5365) were found. Means with identical letters are not
significantly different at the 0.05 level by the Tukey-Kramer procedure. Significant differences exist between species. ACRU=Acer rubrum;
LIST=Liquidambar styraciflua; LITU=Liriodendron tulipifera; NYSY=Nyssa sylvatica; PRSE=Prunus serotina; QUAL=Quercus alba;
QUBI=Quercus bicolor; QUFA=Quercus falcata; QULY=Quercus lyrata; QUPH=Quercus phellos; QURU=Quercus rubra.
51
Ph
osp
ho
rus
(%
)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
Ni
tro
ge
n(
%)
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Day
11-6
-02
11-1
5-02
11-2
0-02
12-2
3-02
Ca
rbo
n(
%)
47
48
49
50
51
52
53
Agricultural
Site
Reference
Site
(a)
(b)
(c)
Figure 3-7. Change in nutrient concentrations in tree litter over time in 2002 for (a) nitrogen, (b)
phosphorus, and (c) carbon. No significant differences exist within sites between dates or between
sites within dates at the 0.05 level.
52
Ph
osp
ho
rus
(%
)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
Ni
tro
ge
n(
%)
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Day
10-1
0-03
10-1
5-03
10-2
2-03
10-2
9-0311-5-03
11-1
4-03
Ca
rbo
n(
%)
47
48
49
50
51
52
53
Agricultural
Site
Reference
Site
(a)
(b)
(c)
a*
b* b*
ab*
aaa
a
a*
b b b*
aaaa
b
ab ab* ab
a
a
a
a
a
bc
a
c
b b
aa
a
ab
a
a*
Figure 3-8. Change in nutrient concentrations in tree litter over time in 2003 for (a) nitrogen, (b)
phosphorus, and (c) carbon. Means with identical letters are not significantly different between dates
within site at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists within dates between sites at the 0.05 level. Significant
differences were observed over time at the agricultural site but not at the reference site and between
sites on some days.
53
N:P ratios between species at both sites and between sites for Symplocarpus foetidus
existed (Figure 3-12). Also, C:P (Figure 3-13) and C:N (Figure 3-14) ratios differed by
species at the agricultural and reference sites.
In 2003, the concentration of nitrogen differed by species and site (Figure 3-15)
while carbon concentrations (Figure 3-16) and the C:N ratios (Figure 3-17) varied by
species only. The C:P ratios differed by species at the agricultural site, but not at the
reference site (Figure 3-18). The phosphorus concentrations did not differ by species
within sites but Leersia oryzoides and Symplocarpus foetidus contained significantly
higher concentrations at the agricultural site than at the reference site (Figure 3-19).
However, the N:P ratios showed no differences between species or site (Figure 3-20).
DISCUSSION
A higher concentration of nitrogen was found in the litter from the agricultural
site than the litter from the reference site during the 2003 season. This may be due to the
elevated amount of nitrate entering the agricultural site from the surrounding fields.
These results agree with other studies. Boerner (1984) studied fertile and infertile forests
and determined that the nitrogen concentrations in litter were higher at the fertile site than
at the infertile site. This elevation in nitrogen may help promote nitrogen mineralization
at the agricultural site by stimulating soil microbes involved in the breakdown the
nitrogen. However, phosphorus concentrations were significantly higher at the
agricultural site than at the reference site in 2003. As stated earlier, the concentration of
phosphate in shallow groundwater samples was higher at the agricultural site than at the
reference site. Thus, more phosphorus is probably available for plant uptake at the
enriched site. Moreover, more carbon was present in the litter from the agricultural site
Agricultural Site Reference Site Agricultural Site Reference SiteParameter
2002 2002 P 2003 2003 P
Biomass
(kg/ha) 1006.0 ? 393.60 2111.9 ? 603.93 0.2024 2322.5 ? 932.61 2573.0 ? 413.40 0.81
N (%) 2.4 ? 0.16 2.3 ? 0.07 0.7857 3.1 ? 0.19 2.8 ? 0.17 0.16
P (%) 0.3 ? 0.03 0.2 ? 0.02 0.1098 0.3 ? 0.03 0.3 ? 0.02 0.11
C (%) 40.3 ? 0.33 40.6 ? 0.38 0.6615 41.7 ? 0.92 36.6 ? 2.20 0.057
N (kg/ha) 26.2 ? 11.10 49.7 ? 15.58 0.3181 80.0 ? 36.84 72.0 ? 14.04 0.84
P (kg/ha) 6.1 ? 1.13 14.2 ? 0.79 0.6510 6.5 ? 2.34 5.8 ? 1.38 0.81
C (kg/ha) 406.0 ? 160.60 859.1 ? 252.98 0.2022 990.8 ? 413.46 946.2 ? 171.30 0.92
N?NUE 42.5 ? 3.50 42.9 ? 1.29 0.9561 32.4 ? 2.18 36.9 ? 2.57 0.21
P?NUE     338.4 ? 37.19 492.3 ? 51.13 0.0767      377.7 ? 38.43 494.9 ? 58.64 0.13
N:P Ratio 8.0 ? 0.53 11.5 ? 31.08 0.0249 11.7 ? 1.06 13.3 ? 0.99 0.31
C:N Ratio 17.1 ? 1.42 17.4 ? 0.36 0.9211 13.5 ? 0.73 13.2 ? 0.24 0.78
C:P Ratio 136.6 ? 15.34 199.9 ? 22.60 0.0789 156.8 ? 14.63 175.6 ? 12.72 0.35
Table 3-2. Biomass, nutrient concentrations, nutrient loads, nutrient use efficiencies, and element ratios for herbaceous vegetation at both sites in 2002
and 2003. Significance values were determined using t-test analyses.
Species
BOC
Y
CAI
N
CIA
R
IMC
A JUEF
LEO
R
MIV
I
OSC
I
PEV
I
PIPU POAR POPU SALA SMRO SYFO UID
Ni
tro
ge
n(
%)
0
1
2
3
4
5
Ni
tro
ge
n(
%)
0
1
2
3
4
5
(a)
(b)
a ab
abab
b ab bab ab
b b
b
b
abab
aabc
abc bc
c
Figure 3-9. Differences in nitrogen concentrations in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site. Means
with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. There were no significant differences
between sites within species at the 0.05 level using the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica; CAIN=Carex intumescens;
CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium vimineum; OSCI=Osmunda
cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum; SALA=Sagittaria latifolia;
SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CIA
R
DUI
N
IMC
A JUEF
LEO
R
MIV
I
OSC
I
PEV
I
PIPU
Poac
eae
sp. POAR POPU SALA
SMR
O
SYF
O UID
Ph
osp
ho
rus
(%
)
0.0
0.2
0.4
0.6
0.8
Ph
osp
ho
rus
(%
)
0.0
0.2
0.4
0.6
0.8
(a)
(b)
bcd abc*
abcd
abc
ab*
cd*abcd
abc abcd
abcd
cd
abc
d
aa aa a a a
abcd
abcd
a
a a
a
Figure 3-10. Differences in phosphorus concentrations in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site.
Means with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*)
beside them indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria
cylindrica; CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides;
MIVI=Microstegium vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium;
POPU=Polygonum punctatum; SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI
PAQ
U PIPU
POA
R
POP
U
SCL
A
SMR
O
SYF
O UID
Ca
rbo
n(
%)
0
10
20
30
40
50
60
Ca
rbo
n(
%)
0
10
20
30
40
50
60
(a)
(b)
ab
a
b ab
a
c
ab abab ab
abab ab
ab
ab
ab
b
a
ab abab ab ab ab ab
ab
ab ab
Figure 3-11. Differences in carbon concentrations in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site. Means
with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CIA
R
IMC
A JUEF
LEO
R
MIV
I
OSC
I
PEV
I
PIPU POAR SALA SMRO SYFO UID
N:
P
0
5
10
15
20
N:
P
0
5
10
15
20
(a)
(b)
ab ab ab
a*
b
ab ab
ab
ab
bab
ab
b
ab
ab
ab
a
ab
ab
Figure 3-12. Differences in N:P ratios in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CIA
R
IMC
A JUEF
LEO
R
MIV
I
OSC
I
PEV
I
PIPU POAR SALA SMRO SYFO UID
C:
P
0
100
200
300
400
500
600
C:
P
0
100
200
300
400
500
600
(a)
(b)
a
b
bcbc
bc bcbc bc bcc bc bc
a
bbcbb
bb
Figure 3-13. Differences in C:P ratios in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CIA
R
IMC
A JUEF
LEO
R
MIV
I
OSC
I
PEV
I
PIPU POAR POPU SALA SMRO SYFO UID
C:
N
0
20
40
60
80
C:
N
0
20
40
60
80
(a)
(b)
a
abc abcab abc
bcabc c abcabc abc abc
a
cbc
bbc bc
bc
abc
Figure 3-14. Differences in C:P ratios in herbaceous vegetation by species in 2002 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Ni
tro
ge
n(
%)
0
1
2
3
4
5
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI
PAQ
U PIPU
POA
R
POP
U
SCL
A
SMR
O
SYF
O UID
Ni
tro
ge
n(
%)
0
1
2
3
4
5
(a)
(b)
a
ab
bc
ababc* ab*
bc
abc ab
abc abc abc
bc
bc
ab
aabc
abc
c
abc c
abc abc
abcabc
Figure 3-15. Differences in nitrogen concentrations in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site. Means
with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI
PAQ
U PIPU
POA
R
POP
U
SCL
A
SMR
O
SYF
O UID
Ca
rbo
n(
%)
0
10
20
30
40
50
60
Ca
rbo
n(
%)
0
10
20
30
40
50
60
(a)
(b)
ab
a
b ab
a
c
ab abab ab
abab ab
ab
ab
ab
b
a
ab abab ab ab ab ab
ab
ab ab
Figure 3-16. Differences in carbon concentrations in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site. Means
with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI
PAQ
U PIPU
POA
R
POP
U
SCL
A
SMR
O
SYF
O UID
C:
N
0
20
40
60
C:
N
0
20
40
60 (a)
(b)
a
bc
c
ab
bc bc bc bc
abc abc
bc
c c bc
bbc
bc
a
abc
bc bc
bc bc
c c c
Figure 3-17. Differences in C:N ratios in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI PIPU POAR POPU SCLA
SMR
O
SYF
O UID
C:
P
0
100
200
300
400
500
600
C:
P
0
100
200
300
400
500
600
(a)
(b)
b
ab abab ab ab
a
ab
ab
ab
ab
b
a
aa
a
a a
a
b
ab
ab
a
a
Figure 3-18. Differences in C:P ratios in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R
DIC
L DIVI
FRV
I
GRA
SSIMCAJUTELEOR MIVI OSCI PIPUPOARPOPEPOPUPOSASCLA
SMR
O
SYF
O UID
Ph
osp
ho
rus
(%
)
0.0
0.2
0.4
0.6
0.8
Ph
osp
ho
rus
(%
)
0.0
0.2
0.4
0.6
0.8
(a)
(b)
* *
Figure 3-19. Differences in phosphorus concentrations in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site.
Means with identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*)
beside them indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria
cylindrica; CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides;
MIVI=Microstegium vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium;
POPU=Polygonum punctatum; SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
Species
BOC
Y
CAI
N
CAL
U
CIA
R DIVI
FRV
I
GRA
SS IMCA LEOR MIVI OSCI PIPU POAR POPU SCLA
SMR
O
SYF
O UID
N:
P
0
5
10
15
20
N:
P
0
5
10
15
20
(a)
(b)
Figure 3-20. Differences in N:P ratios in herbaceous vegetation by species in 2003 at the (a) agricultural site and (b) reference site. Means with
identical letters are not significantly different within sites at the 0.05 level by the Tukey-Kramer procedure. Letters with asterisks (*) beside them
indicate a significant difference exists between sites within species at the 0.05 level by the Tukey-Kramer procedure. BOCY= Boehemeria cylindrica;
CAIN=Carex intumescens; CIAR=Cinna arifolium; IMCA=Impatiens capensis; JUEF=Juncus effusus; LEOR=Leersia oryzoides; MIVI=Microstegium
vimineum; OSCI=Osmunda cinnamomea; PEVI=Peltandra virginica; PIPU=Pilea pumila; POAR=Polygonum arifolium; POPU=Polygonum punctatum;
SALA=Sagittaria latifolia; SMRO=Smilax rotundifolium; SYFO=Symplocarpus foetidus; UID=unidentified species.
67
than the litter from the reference site in both years. This difference may be due to the
fragmented structure of the agricultural site. Since this site is narrower, more light can
potentially penetrate the canopy. This is most likely causing photosynthesis rates to
increase ultimately increasing carbon production.
The nitrogen NUE was significantly higher at the reference site than at the
agricultural site in 2003. Once again this difference between years could be due to the
drought or lack of monitoring plot replication in 2002. It has been found that higher
NUEs exist in forests with lower aboveground nitrogen circulation (Vitousek 1982). Our
results appear to support their findings. Nutrient rich sites may have higher
concentrations in leaf litter than non-enriched sites do. However, since nitrogen is more
readily available in these enriched sites, the trees may not have to resorb higher
concentrations of nitrogen from the leaves prior to abscission. Another possible reason
for this discrepancy in NUEs between sites is that the nitrogen resorption rate may not
increase in response to higher nitrogen concentrations. This is extremely important in the
cycling of nutrients through ecosystems because if resorption rates do not increase in
response to higher nutrient concentrations, then the potential for loss of nutrients via
throughfall is greater. These results suggest that enriched sites may have a higher amount
of nutrients returned to the system via litterfall than non-enriched sites.
N:P ratios are important because they serve as indicators of microbial processes.
In 2002, the N:P ratio was higher at the agricultural site than the reference site due to the
elevated nitrogen in the litter at the agricultural site. Both sites had N:P ratios greater
than 12 in both years. Also, N:P ratios greater than 10 may actually cause phosphorus to
be immobilized (Lockaby 1999). The C:N ratios were not significantly different in 2002.
68
Nevertheless in 2003 the C:N ratio was significantly higher at the reference site than at
the agricultural site. This result occurs because the agricultural site has higher
concentrations of both nitrogen and carbon. The C:N ratios at both sites were greater
than 40 in both years. These C:N ratios are higher than those found by Bischoff et al.
(2001). C:N ratios greater than 20 to 30:1 will cause the immobilization of some nitrogen
since the microorganisms will preferentially use more easily attainable nitrogen in the
soils (Barbour et al. 1999). This is relevant since areas with high C:N ratios may release
nitrogen slowly during decomposition causing the net nitrogen mineralization to be small.
C:P ratios were not significantly different between sites in either year.
Not finding a significant difference in litter biomass was definitely an unexpected
result that is probably due to the characteristic differences observed at each site. Unlike
the reference site, the agricultural site is fragmented in nature and winds may actually
cause the export of litter from the system. Thus wind could actually limit the amount of
biomass available for collection at the agricultural site. In addition, no differences in the
nutrient loads were detected. This result seems counterintuitive at first. However, it is
probably a reflection of the higher biomass at the reference site. The lower litter input at
the agricultural site may be offsetting the nutrient loading rates due to elevated
concentrations of nitrogen, phosphorus, and carbon. Bischoff et al. (2001) found that the
average nitrogen load in a deciduous forested wetland was 20 kg/ha while the average
carbon load was 661 kg/ha. When simply considering the year with normal climatic
patterns (i.e. 2003), my nitrogen values (22.54?2.2 kg/ha at the agricultural site and
24.4? 2.3 kg/ha at the reference site) are consistent with Bischoff?s estimate.  However, 
my carbon loads (1010.61?108.2 kg/ha at the agricultural site and 1222.10?98.5
69
kg/ha) are higher than those found by Bischoff (2001). Yet, the average litter biomass of
their site was also lower (1302 kg/ha) than mine (1981.75?216.9 at the agricultural site
and 2450.1?197.9 at the reference site) (Bischoff et al. 2001).
The 2002 data was typified by an overall lack of differentiation between the two
sites for both tree leaf litter and herbaceous vegetation. The carbon concentrations and
NUE in leaf litter and N:P ratios in herbaceous vegetation were the only overall
differences found between sites. This may be due to several factors. A severe drought
occurred from 2001-2002. By March of 2002 the Maryland Department of the
Environment (MDE) had declared central Maryland, which includes both study sites, to
be in an imminent emergency (MDE 2005). By August of that same year Maryland was
experiencing an ?exceptional drought,? which means that crops would most likely be 
adversely impacted, fire risks were high, and water shortages were likely imminent.
Since the agricultural site is strongly influenced by the contribution of groundwater, a
decrease in groundwater (and surface water) levels may have limited the flow of nitrogen
into the system and consequently its availability for plant uptake. Furthermore, I may not
have detected a difference in 2002 because I only had two monitoring plots at the
reference site compared to six at the agricultural site. This lack of replication may have
limited my ability to detect any existed differences between sites.
There were significant differences between species within both sites but
insignificant differences between sites by species for tree litter. The concentrations of
nitrogen within the species are similar to those found in other studies. It appears that in
general the Quercus species (oaks) had higher nutrient concentrations than the other
species. These results coincide with those of Killingbeck and Tainsh (2002). They found
70
that Acer rubrum had more proficient resorption of nitrogen than Quercus alba did
(Killingbeck and Tainsh 2002). This result suggests that oaks may return higher
concentrations of nitrogen to the system than the other species. Oak leaves are generally
larger and thicker than the leaves of the other species studied here. It is easy to cite this
discrepancy as a possible cause for these results. However, others have found that leaf
size and thickness does not affect the resorption of nitrogen and phosphorus in Acer
rubrum and Quercus alba leaves (Killingbeck and Tainsh 2002). The discrepancy may
be due to differences inherent in the species. The concentrations in Quercus alba and
Quercus rubra are comparative to those found in other studies (Killingbeck and Tainsh
2002; Killingbeck 1996; Boerner 1984). However, the concentrations in Acer rubrum
and Nyssa sylvatica are higher at both sites in this study than in others (Killingbeck and
Tainsh 2002; Killingbeck 1996; Boerner 1984).
Killingbeck (1996) proposed nitrogen resorption levels for deciduous tree species.
He suggested that values below 0.7 % reflected complete resorption, those above 1.0 %
reflected incomplete resorption, and those in between these values reflected intermediate
resorption. Comparisons based on this classification revealed tha no species at either site
resorbed nitrogen completely. There were more species at the agricultural site that would
be classified as having incomplete resorption than there were at the reference site. In
fact, over 77% of the species at the agricultural site resorbed nitrogen incompletely
whereas over 77% of the species at the reference site resorbed nitrogen at an intermediate
level.
The concentrations significantly differed depending on the species but did not
differ within species between sites. I expected to find differences within species. The
71
Quercus rubra leaves at the agricultural site contained significantly higher concentrations
of phosphorus than did Acer rubrum, Quercus alba, Quercus falcata, and Quercus
phellos. Killingbeck (1996) found phosphorus concentrations in Quercus rubra leaves
were 0.07 % and those in Nyssa sylvatica were 0.04 %. In this study, the phosphorus
concentrations were slightly higher for both species?0.098 at the agricultural site and
0.082 at the reference site for Quercus rubra and 0.071 at the agricultural site and 0.091
at the reference site for Nyssa sylvatica. Acer rubrum and Quercus alba concentrations
were comparable to those found in other studies (Killingbeck and Tainsh 2002;
Killingbeck 1996).
Killingbeck (1996) also proposed phosphorus resorption levels for deciduous
trees. He asserted that complete resorption occurred below 0.05% while incomplete
resorption was above 0.08%. Intermediate resorption occurred in between these two
values. Once again no species completely resorbed the nutrient. Moreover, almost 77%
of the species at the agricultural site resorbed phosphorus incompletely while slightly
over 55% of the species at the reference site resorbed phosphorus at an intermediate
level.
This time some differences in carbon were observed between species within site
and between sites by species. Inconsistencies between species within sites could be due
to differences in the ability of the trees to harvest sunlight for photosynthesis. Overall,
the oak species have higher carbon concentrations in their leaves than do the other
species. The oak leaves are larger than the other species thereby allowing for a greater
surface area to capture light.
72
The carbon concentrations were also significantly greater in leaves of Acer
rubrum, Liquidambar styraciflua, and Quercus rubra at the agricultural site than at the
reference site. Carbon is closely linked to nitrogen (Barbour et al. 1999). Acquiring
nitrogen requires energy supplied by carbon metabolism. In return, biomass production
is often limited by nitrogen availability (Barbour et al. 1999). Since nitrogen is more
readily available to the trees at the agricultural site, more biomass may be produced
causing higher carbon concentration in the litter.
There was only one difference found between species for the N:P ratios. Quercus
falcata leaves had significantly higher N:P ratios than did Quercus rubra. Boerner
(1984) determined nitrogen and phosphorus concentrations on a weight basis in Acer
rubrum and Quercus alba leaves in a fertile and infertile site. Because he did not present
N:P ratios per se, I took the nitrogen and phosphorus concentrations that he presented and
calculated N:P ratios for each of his study sites. I then averaged them together so that I
could determine if the values obtained in my study were comparable to others. Based on
data presented by Boerner (1984), the average N:P ratio for Acer rubrum was 11.5 and
for Quercus alba was 12.1. These values are lower than my values of 14.9 and 17.5 for
Acer rubrum and Quercus alba, respectively.
Although differences between species existed at both sites, there were more
differences at the agricultural site than at the reference site. The nitrogen enrichment at
the agricultural site may be causing some trees to take up more nitrogen and increase
carbon production (i.e. biomass) thereby altering the natural C:N ratios. This is relevant
since high C:N ratios often cause decomposers to be limited by nitrogen (Vitousek 1982).
Only one difference was found between sites by species. Quercus falcata leaves at the
73
agricultural site had significantly higher C:N ratios than those at the reference site.
The leaves of Acer rubrum and Quercus falcata had significantly higher C:P
ratios than did Quercus rubra at the agricultural site. Despite this difference at the
agricultural site, there were no differences between species at the reference site.
Similarly no significant differences existed between sites within species. However, the
C:P ratios were higher in all but two species at the agricultural site than at the reference
site. This result is attributable to differences in carbon more so than differences in
phosphorus.
In 2003 the leaves fell within a six week period. A decreasing trend in nitrogen
concentrations was observed in 2003 from October 10 through November 5 at the
agricultural site was expected since the nutrients were being resorbed by the trees.
Woodwell (1974) studied tree litter of Quercus alba, Quercus coccinea, and Pinus rigida
in a New York forest. His results were consistent with mine, for he found that nitrogen
concentrations in tree leaf litter decreased throughout the period of senescence
(Woodwell 1974). This suggests that leaves remaining attached to trees for an extended
period of time had a greater concentration of nitrogen resorbed by the woody tissues.
The increase in concentration that occurred on the last date was apparently caused by a
frost with a low temperature of 24?F lasting 3 days that happened just prior to this
collection date (NOAA). The leaves froze, essentially trapping the nitrogen in the leaf
and preventing the tree from withdrawing it. An increase was not observed on the same
date at the reference site possibly because, unlike the fragmented agricultural site, the
contiguous span of forest provides a buffer against low temperatures and high winds.
Significant differences existed between sites for nitrogen on all dates except on October
74
29 and November 5, 2003. Boerner (1984) also found that trees growing in fertile
environments had higher concentrations of nitrogen in tree litter than those in infertile
environments. As just previously mentioned, the frost may explain the discrepancy at the
end of the year. The difference at the beginning of the year is probably a reflection of the
excess nitrogen at the agricultural site.
In addition, the concentration of phosphorus did decrease over time in 2003
except for on the last collection date. Woodwell (1974) found similar results throughout
the period of senesce in New York. As previously mentioned, a frost occurred just prior
to that date and provides a possible explanation for the observed, yet insignificant,
difference. Significant differences in phosphorus between sites were detected at the start
and end of the season only. The difference at the beginning of the season is attributed to
variations in phosphorus availability between the sites. The discrepancy at the end of the
season is probably due to the frost.
In 2003 carbon varied significantly within and between sites by date. The
observed increase in carbon at the agricultural site is expected. A significant difference
between sites in the middle of the season was also seen. This result could be due to an
abundance of structural carbon that remains in the leaf later in the season. As nitrogen
and phosphorus are resorbed the proportion of carbon left in the leaf increases. This is
important since the amount of carbon being resorbed by senescing leaves has been found
to influence the efficiency of nutrient resorption (Chapin and Moilanen 1991).
Recall that in 2002 only two plots had litter traps place within them at the
reference site. There may have been insufficient replication at this site to detect
differences over time. There was also no difference between sites in 2002. As stated
75
earlier, this result may be due to climatic factors. A significant drought occurred for
much of 2002. The lack of precipitation may have caused the trees to drop their leaves
earlier in the season and the period of senescence may have been shortened thereby
altering the normal resorption rates. The majority of the litter was collected within a
three week period during 2002. Del Arco et al. (1991) found that when the annual
rainfall influenced retranslocation efficiency. He concluded that the higher the annual
rainfall, the higher the retranslocation efficiency (Del Arco et al. 1991). Although I did
not measure retranslocation efficiency, the influence of drought may help to explain the
lack of decrease in nutrients during the period of senescence.
Furthermore, overall comparisons of the herbaceous vegetation revealed only two
significant differences between sites: in 2002, the N:P ratio was significantly higher at
the reference site than at the agricultural site and in 2003 the concentration of carbon was
marginally significantly higher at the agricultural site than at the reference site. The
difference in N:P ratios can be explained by the drought of 2002. The agricultural site is
largely influenced by groundwater inputs (Angier et al. 2002) so the drought limited the
ability of the groundwater to transport nitrogen into the agricultural site. The plants
found at the reference site were probably already adapted to low nitrogen availability but
those at the agricultural site probably were not. The higher carbon concentrations at the
agricultural site in 2003 may be attributed to the higher light levels at the agricultural
causing increasing photosynthesis rates. There were some differences in nutrient
concentrations and element ratios within site by herbaceous species during both years and
at both sites. This difference between species has been found in other studies (Kao et al.
2003). Any differences that occurred between sites by species were partly due to
76
Symplocarpus foetidus. Thus, this species may be able to take up excess nutrients in
enriched sites.
CONCLUSIONS
A lower number of significant differences were found in 2002 than in 2003. This
could be due to a lower number of repetitions at the reference site. The two plots found
there may not have been sufficient to detect differences. Also, since 2002 was a drought
year, the behavior of nutrient cycling may have been different than in a year when normal
climatic patterns occur like in 2003 possibly due to plant stress caused by a lack of water.
Moreover, the influence of groundwater appears to be especially important for the
vegetation at the agricultural site. Because the plants in this system appear to rely heavily
on the import of nutrients via groundwater, when the levels dropped, the availability of
nutrients probably did too. In addition, there were more differences detected between
sites for the tree litter than for the herbaceous vegetation. This may also be attributed to
the role of groundwater. The levels of nitrate in the groundwater are lowest in the
shallow groundwater and increase with depth. Thus, the trees, which roots that extend
deep into the soil profile, may potentially take up higher concentrations of nutrients. The
herbaceous vegetation, on the other hand, has shallow roots that are exposed to lower
concentrations of nutrients. Therefore, they probably are not taking up concentrations of
nutrients like the trees.
During the 2003 season, the nitrogen, phosphorus, and carbon concentrations in
tree litter were all significantly greater at the agricultural site than at the reference site.
Moreover, the NUEs of nitrogen and phosphorus were both significantly lower at the
agricultural site than at the reference site. Based on these results the trees at the
77
agricultural site return higher concentrations of nutrients to the system via litterfall. The
observed decrease in nutrient concentrations during the 2003 senescent period at the
agricultural site reveals that the nutrient cycles at the site may be changing in response to
the excess nitrogen and phosphorus. Despite these decreasing concentrations, nitrogen
and phosphorus significantly increased in the agricultural litter when a frost occurred.
The fragmented nature of these agricultural buffers may make more susceptible to frost
damage and could temporarily change nutrient patterns within these systems.
78
Chapter 4: The Effects of Nutrient Enrichment on the
Herbaceous Vegetation of a Forested Riparian Wetland
INTRODUCTION
Surface and ground water exiting agricultural fields often contain high levels of
nitrogen and phosphorus. These nutrients then enter nearby ecosystems where they
interact with the vegetation. These additional nutrients can cause changes in the plant
community to occur. Community level changes, such as a decrease in species diversity,
species richness and species evenness have been documented in communities enhanced
by nutrients (Roberts and Gilliam 1995; Bedford et al. 1999; Joyce 2001). Furthermore,
species level changes, including changes in species competition and plant uptake of
nutrients, can also occur due to elevated concentrations of nutrients (Roberts and Gilliam
1995). The addition of the limiting nutrient(s) (often nitrogen, phosphorus, or a
combination of the two) by definition results in an increase in biomass production while
the addition of non-limiting nutrients produces little if any effect (Silvan et al. 2004).
It is important to study the influence of nutrients on the vegetation found in
natural settings so that we can understand how excess nutrients from human activities
alter plant vegetation and biodiversity. Only by fertilizing these plants can we gain
insight into the direct effects of the agricultural nutrients on the individual plant species
and the ecosystem as a whole. Thus, we need to identify those species that are likely to
be negatively impacted by increased nutrients.
It has been argued that the N:P ratios of plants can reveal whether plants are
limited by nitrogen, phosphorus, or a co-limited by both nutrients. There are at least two
79
main schools of thought on the relationship of N:P ratios and nutrient limitation.
Koerselman and Meuleman (1996) reviewed the results of forty experiments involving
the fertilization of herbaceous vegetation was fertilized in natural wetland settings in
Europe. They suggested that N:P ratios less than 14 reflect nitrogen limitation while
those greater than 16 reflect phosphorus limitation. N:P values in between 14 and 16
represent co-limitation by nitrogen and phosphorus.
The resource ratio model, on the other hand, provides a different view. It predicts
that if a species is co-limited by nitrogen and phosphorus, the addition of nitrogen will
cause that species to compete for phosphorus since it would be in more limited in
quantity (Tilman 1985). The species that can increase phosphorus uptake could then
outcompete those species that could not. The opposite outcome would occur if
phosphorus was suddenly added to the system and nitrogen became limited. This model
also predicts that the concentrations of nutrients in plant biomass are a reflection of the
competitiveness of the plant. Those plants that are strong competitors would contain low
concentrations of the nutrient while those that are weak competitors would contain high
concentrations.
I examined responses of herbaceous vegetation to excess nitrogen and phosphorus
in a fertilization experiment in a forested riparian wetland. Percent cover and frequency
were determined and biomass and nitrogen, phosphorus, and carbon concentrations in
plant tissue were measured. Nutrient loads, nutrient use efficiencies (the plant biomass
produced for every unit of nutrient mass), and element ratios were also calculated.
Comparisons of these parameters between treatments were made first by averaging across
species to determine community level trends and then by inspecting by species to
80
determine trends at the species level. Four species that were present in every plot were
examined?Impatiens capensis (Meerb.), Microstegium vimineum (Trin.) A. Capus,
Polygonum punctatum Ell., and Symplocarpus foetidus L. Salisb. ex. Nutt. Microstegium
vimineum was the only non-native species present. This C4 grass originated in Japan and
has become naturalized throughout the eastern and southeastern United States (Claridge
and Franklin 2002).
I hypothesized that the community as a whole would be limited by nitrogen since
many freshwater wetlands have been shown to be nitrogen limited (Bedford et al. 1999).
I also expected that resource limitation would vary by species. Furthermore, I anticipated
that species would differ in nutrient concentrations, NUEs, and element ratios since there
is often variability in nutrient concentrations between species.
METHODS
The fertilization experiment was conducted at the reference site only. No
experiments were conducted at the agricultural site. The direct addition of nutrients into
the agricultural system may have influenced studies that other researchers were
performing simultaneously (like the monitoring of nitrogen concentrations in
groundwater and surface water). Moreover, it was unknown whether any fertilization
effects would be pronounced enough to detect since the plants at this site have been
exposed to elevated concentrations of nutrients for decades.
Twenty 2 x 1 m plots were randomly distributed throughout the reference site and
marked with PVC poles. Nitrogen and phosphorus were applied in a 2 x 2 factorial
arrangement, resulting in four treatment combinations: (1) the addition of nitrogen, (2)
the addition of phosphorus, (3) the addition of nitrogen and phosphorus, and (4) no
81
additions. Each treatment combination was randomly applied to five of the 2 x 1 m plots.
Fertilizer in the form of urea (N-P-K 46-0-0) for nitrogen and triple superphosphate (N-P-
K 0-46-0) for phosphorus were applied. Since the study area is frequently flooded, I
decided that the fertilizer should be applied below the ground surface in the rooting zone
instead of simply being applied on the ground surface. By digging holes I hoped to
ensure that the fertilizer would not be exported out of the system by flood waters.
I cut eight holes at equal intervals (two rows of four) into a piece of plastic that
was 1 x 2 m in area. The plastic was then used as a guide and laid on the soil surface
within each plot. Eight holes were then dug within each plot using a soil corer. Plots that
did not receive additional nitrogen or phosphorus likewise had holes dug to maintain
consistency and to ensure that possible effects due to the holes would be present in all
plots. Within each of the holes 150 g of the assigned treatment was applied. Thus,
nitrogen treated plots received a total of 1200 grams of nitrogen; likewise, phosphorus
treated plots received a total of 1200 grams of phosphorus. Plots fertilized with both
nitrogen and phosphorus received 150 g of each nutrient within each hole resulting in a
total of 2400 g of fertilizer (1200 g of nitrogen and 1200 g of phosphorus). The holes
were then resealed with soil.
The percent cover and frequency of each species in all plots were determined in
the end of May. A sampling frame quadrat subdivided into 32 cells was placed over each
2 x 1 m plot to determine the frequency of occurrence of individual species (i.e., the
number of cells it occurred in divided by 32 multiplied by 100%). Percent cover was
determined for the entire 2 x 1 m quadrat. The percent cover is the area of ground
surface that each plant occupies if projected vertically onto the surface within the quadrat
82
expressed as a percentage. Percent cover values can be greater than 100% since plants
often overlap each other. The cover values used were: <0.1%, 0.1% increments from 0.1-
1%, 1% increments from 1-10%, and 5% increments above 10%.
The aboveground biomass within each 2 x 1 m plot was harvested in June,
separated by species while harvesting in the field, dried at 60?C until constant weight was
achieved, weighed, homogenized using a Wiley Mill, and analyzed for nitrogen,
phosphorus, and carbon. Total nitrogen and carbon were determined using a LECO Tru-
Spec CN Analyzer (St. Louis, Missouri). Total phosphorus concentrations were
determined using a modified Kjeldahl digestion method (instead of using 3.5 mL of
sulfuric acid as called for in this method, I used 5 mL). Digests were then run on a Flow
Injection Analyzer (Lachat QuikChem 8000, Milwaukee, Wisconsin).
Phosphorus concentrations were originally given in units of parts per million
(ppm=?g/mL) while nitrogen and carbon were given in units of percent. The phosphorus
concentrations were therefore converted to units of percent so that all data would be in
the same units. Nutrient loads were calculated by dividing the nutrient mass (in units of
kilograms) by the total area of each plot in hectares). Nutrient use efficiencies were
calculated by dividing the biomass in units of grams by the nutrient mass (which were
converted into units of grams). Element ratios were also calculated using nutrient
contents in units of mass. All statistical analyses on the data were then performed using
SAS (v. 8 for Windows, SAS Institute, Cary, NC, USA). The data were examined by
averaging across species to investigate general plant community level trends due to
nitrogen and phosphorus fertilization and by averaging within species by treatment to
determine possible differences between species.
83
RESULTS
When considering overall trends due to fertilization, the addition of nitrogen and
phosphorus did increase the concentration of nitrogen slightly but did not alter it
significantly (Figure 4-1a; Table 4-1). Although the concentration of phosphorus was not
affected by the addition of nitrogen, it was significantly increased when phosphorus was
added (Figure 4-1b; Table 4-1). Carbon concentrations were unaffected by the addition
of nitrogen or phosphorus (Figure 4-1c; Table 4-1).
In addition, the application of phosphorus decreased the N:P ratio significantly
while that of nitrogen did not (Figure 4-2a; Table 4-1). However, neither the C:N (Figure
4-2b; Table 4-1) nor the C:P (Figure 4-2c; Table 4-1) ratios were influenced significantly
by the addition of nitrogen or phosphorus.
Despite the lack of significant differences of nutrient loads due to fertilization
treatment, the phosphorus load was slightly increased when nitrogen was added but
decreased when phosphorus was added (Figure 4-3a; Table 4-1). The application of
nitrogen did increase the nitrogen load but not significantly while the addition of
phosphorus, actually decreased the nitrogen load (Figure 4-3b; Table 4-1). Moreover,
carbon followed a similar trend as the other two nutrients studied. The addition of
nitrogen increased the carbon load while the addition of phosphorus decreased it (Figure
4-3c; Table 4-1).
The addition of nitrogen had no significant effect on the nitrogen NUE; likewise
phosphorus additions did not alter the NUE of nitrogen (Figure 4-4a; Table 4-1). Also,
the NUE of phosphorus was not affected when nitrogen was added to the system, but it
was significantly decreased when phosphorus was added (Figure 4-4b; Table 4-1).
84
Ph
osp
ho
rus
(%
)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Ni
tro
ge
n(
%)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Nitrogen
Ca
rbo
n(
%)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
*
? + ? +
Phosphorus
(a)
(c)
(b)
Figure 4-1. Nitrogen and phosphorus effects on the concentrations of (a) nitrogen, (b) phosphorus,
and (c) carbon of all herbaceous species. Values with asterisks (*) between them indicate a
significant treatment effect for that nutrient at the 0.05 level. ? = no nutrient added; + = nutrient
added.
85
Parameter Nitrogen Phosphorus Nitrogen*Phosphorus
Phosphorus 0.09(1,16) 4.40(1,16) 0.03(1,16)
Carbon 0.90(1,16) 0.34(1,16) 0.10(1,16)
Nitrogen 0.90(1,16) 0.34(1,16) 0.10(1,16)
C:N 2.77(1,16) 1.92(1,16) 0.27(1,16)
C:P 0.03(1,16) 1.28(1,16) 1.40(1,16)
N:P 2.02(1,16) 5.51(1,16)* 0.28(1,16)
P load 1.59(1,16) 0.97(1,16) 1.20(1,16)
N load 2.61(1,16) 3.00(1,16) 1.10(1,16)
C load 1.92(1,16) 2.63(1,16) 0.83(1,16)
NUE-P 0.30(1,16) 4.16(1,16) 0.08(1,16)
NUE - N 0.17(1,16) 0.09(1,16) 0.11(1,16)
Biomass 1.59(1,16) 1.81(1,16) 0.29(1,16)
Cover 0.30(1,756) 0.47(1,756) 0.13(1,756)
Frequency 0.03(1,756) 0.22(1,756) 0.03(1,756)
Table 4-1. Tests of fixed effects results for all species. Values are the F values and the numbers in
parentheses are the numerator and denominator degrees of freedom, respectively. N is the nitrogen
treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus interaction
effect. An asterisks (*) indicates significance at the 0.05 level. There was a significant phosphorus
treatment effect for the N:P ratio.
86
C:
P
0
20
40
60
80
100
120
140
160
180
200
C:
N
0
5
10
15
20
25
N:
P
0
2
4
6
8
10
12
14
16
18 *
Treatment
Nitrogen
? + ? +
Phosphorus
Figure 4-2. Nitrogen and phosphorus effects on element ratios of all herbaceous species. Values with
asterisks (*) between them indicate a significant treatment effect for that nutrient at the 0.05 level.
The horizontal lines in the N:P graph represent the cutoff values for nutrient limitation suggested by
Koerselman and Mueleman (1996). Values above 16 represent phosphorus limitation while those
below 14 represent nitrogen limitation. Values between 14 and 16 represent co-limitation by both
nitrogen and phosphorus. The N:P ratio was significantly decreased when phosphorus was added.
? = no nutrient added; + = nutrient added.
87
P(
kg
/ha
)
0
1
2
3
4
N
(kg
/ha
)
0
10
20
30
40
50
60
C
(kg
/ha
)
0
10
20
30
40
50
60
Nitrogen
? + ? +
Phosphorus
Treatment
(a)
(b)
(c)
Figure 4-3. Nitrogen and phosphorus effects on the nutrient loads of (a) nitrogen, (b) phosphorus,
and(c) carbon for all herbaceous species. There were no significant treatment effects at the 0.05
level.  ? = no nutrient added; + = nutrient added.
88
Ph
osp
ho
rus
N
UE
0
100
200
300
400
500
600
Ni
tro
ge
nN
UE
0
10
20
30
40
50
*
Nitrogen
? + ? +
Phosphorus
(a)
(b)
Treatment
Figure 4-4. Nitrogen and phosphorus effects on the nutrient use efficiencies (NUE is the leaf litter
biomass per unit nutrient content) of (a) nitrogen and (b) phosphorus of all herbaceous species.
Values with asterisks (*) between them indicate a significant treatment effect for that nutrient at the
0.05 level. ? = no nutrient added; + = nutrient added.
89
Neither nitrogen nor phosphorus application affected the overall cover or
frequency of the plants (Table 4-1; Table 4-2). Likewise, the addition of nitrogen or
phosphorus produced no significant differences in biomass (Figure 4-5; Table 4-1).
However, when nitrogen was added, the biomass did increase slightly while the addition
of phosphorus decreased it.
Cover (%) Frequency (%)Effect
Nitrogen Phosphorus Nitrogen Phosphorus
? 2.8 ? 0.6 3.3 ? 0.6 6.2 ? 0.9 5.8 ? 0.9
+ 3.2 ? 0.6 2.7 ? 0.6 5.9 ? 0.9 6.4 ? 0.9
Table 4-2. Nitrogen and phosphorus effects on percent cover and frequency of all herbaceous
species. There were no differences in cover or frequency due to nitrogen or phosphorus. ? = no
nutrient added; + = nutrient added.
Nitrogen
Bi
om
ass
(g
)
0
50
100
150
200
250
? + +?
Phosphorus
Treatment
Figure 4-5. Nitrogen and phosphorus effects on the biomass of all herbaceous species. Values with
asterisks (*) above them indicate a significant difference exists between treatments at the 0.05 level
using tests of fixed effects. No differences in biomass were observed due to nitrogen or phosphorus
additions.  ? = no nutrient added; + = nutrient added.
Differences between the most commonly occurring species were also
investigated. While there were no significant differences in cover due to nitrogen or
90
phosphorus additions in Impatiens capensis (Figure 4-6a; Table 4-3), Microstegium
vimineum (Figure 4-6b; Table 4-4), or Polygonum punctatum (Figure 4-6c; Table 4-5),
there was a significant interaction of nitrogen and phosphorus for Symplocarpus foetidus
(Figure 4-6d; Table 4-6). When phosphorus was added, the percent cover was actually
lower than when no phosphorus was added. The addition of nitrogen alone had the
highest percent cover and the lowest occurred when both nitrogen and phosphorus were
added. There were no significant differences in frequency due to the addition of nitrogen
or phosphorus for any of the species (Figure 4-7 a-d; Tables 4-3, 4-4, 4-5, and 4-6).
Although the phosphorus concentration in Impatiens capensis was not affected
significantly by the addition of nitrogen or phosphorus, it did increase slightly with the
addition of phosphorus (Figure 4-8a; Table 4-3). There was an interaction effect
involving nitrogen and phosphorus, but the concentration of phosphorus in Microstegium
vimineum was always higher when phosphorus was added than when no phosphorus was
added (Figure 4-8b; Table 4-4). The application of nitrogen did not affect the
concentration of phosphorus in Polygonum punctatum, but the application of phosphorus
significantly enhanced the concentration of phosphorus (Figure 4-8c; Table 4-5).
Furthermore, nitrogen and phosphorus additions neither increased nor decreased the
concentration of phosphorus in Symplocarpus foetidus (Figure 4-8d; Table 4-6). There
were no significant differences in nitrogen concentration observed for any of the species
at the 0.05 level using the tests of fixed effects due to the application of nitrogen or
phosphorus (Figure 4-9a-d; Tables 4-3, 4-4, 4-5, and 4-6). However, the application of
nitrogen did slightly increase the concentrations of nitrogen in all species (Figure 4-9a-d;
Tables 4-3, 4-4, 4-5, and 4-6). Carbon concentrations were also examined. There were
91
Impatiens capensis
Co
ve
r(
%)
0
1
2
3
4
5
6
Microstegium vimineum
Co
ve
r(
%)
0
5
10
15
20
25
30
Polygonum punctatum
Co
ve
r(
%)
0.0
0.5
1.0
1.5
2.0
(a) (b)
(c)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Symplocarpus foetidus
Nitrogen Treatment
None Added
Co
ve
r(
%)
0
20
40
60
80
100
(d)
P0
P1
*
Figure 4-6. Nitrogen and phosphorus treatment effects on the percent cover of (a) Impatiens
capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus.
Values with asterisks (*) above them indicate a significant treatment effect for that nutrient at the
0.05 level. Main effects are shown for Impatiens capensis, Microstegium vimineum, and Polygonum
punctatum since the nitrogen*phosphorus interactions were insignificant; nitrogen*phosphorus
interactions are presented for Symplocarpus foetidus since they were significant (P=0.0326). No
significant effects on nitrogen concentration were found when nitrogen or phosphorus was added.  ?
= no nutrient added; + = nutrient added.
92
Parameter Nitrogen Phosphorus Nitrogen*Phosphorus
Biomass 0.24(1,8) 0.05(1,8) 1.23(1,8)
Phosphorus 0.01(1,7) 1.81(1,7) 1.35(1,7)
Carbon 1.14(1,7) 0.68(1,7) 0.03(1,7)
Nitrogen 1.78(1,7) 0.02(1,7) 0.13(1,7)
C:N 1.30(1,7) 0.10(1,7) 0.05(1,7)
C:P 0.000(1,6) 1.51(1,6) 1.60(1,6)
N:P 1.09(1,6) 0.26(1,6) 0.10(1,6)
Cover 2.71(1,16) 0.02(1,16) 0.73(1,16)
Frequency 1.40(1,16) 1.96(1,16) 0.93(1,16)
Table 4-3. Tests of fixed effects results for Impatiens capensis. Values are the F values and the
numbers in parentheses are the numerator and denominator degrees of freedom, respectively. N is
the nitrogen treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus
interaction effect. Neither the interactions nor the main treatment effects were significant.
93
Parameter Nitrogen Phosphorus Nitrogen*Phosphorus
Biomass (kg/ha) 0.69(1,10) 0.31(1,10) 2.31(1,10)
Phosphorus (%) 6.64(1,9)* 38.81(1,9)** 9.66(1,9)*
Carbon (%) 0.05(1,9) 0.43(1,9) 1.73(1,9)
Nitrogen (%) 0.02(1,9) 1.21(1,9) 0.33(1,9)
C:N 0.10(1,9) 0.39(1,9) 0.04(1,9)
C:P 6.65(1,8)* 24.15(1,8)** 3.68(1,8)
N:P 8.68(1,8)* 21.82(1,8)** 1.93(1,8)
Cover 0.16(1,16) 0.08(1,16) 0.09(1,16)
Frequency 0.09(1,16) 0.54(1,16) 0.06(1,16)
Table 4-4. Tests of fixed effects for Microstegium vimineum. Values are the F values and the
numbers in parentheses are the numerator and denominator degrees of freedom, respectively. N is
the nitrogen treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus
interaction effect. An asterisks (*) indicates the significance level as follows: * = <0.05; **= <0.01.
94
Parameter Nitrogen Phosphorus Nitrogen*Phosphorus
Biomass (kg/ha) 2.35(1,2) 0.90(1,2) 47.13(1,2)*
Phosphorus (%) 1.38(1,2) 19.26(1,2)* 0.28(1,2)
Carbon (%) 0.03(1,2) 0.19(1,2) 0.44(1,2)
Nitrogen (%) 1.79(1,2) 0.14(1,2) 1.25(1,2)
C:N 1.49(1,2) 0.25(1,2) 0.03(1,2)
C:P 0.69(1,2) 6.89(1,2) 0.48(1,2)
N:P 23.74(1,2)* 81.45(1,2)* 3.25(1,2)
Cover 0.03(1,16) 0.19(1,16) 0.66(1,16)
Frequency 0.24(1,16) 1.07(1,16) 0.15(1,16)
Table 4-5. Tests of fixed effects for Polygonum punctatum. Values are the F values and the numbers
in parentheses are the numerator and denominator degrees of freedom, respectively. N is the
nitrogen treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus
interaction effect. An asterisks (*) indicates significance at the 0.05 level.
95
Parameter Nitrogen Phosphorus Nitrogen*Phosphorus
Biomass (kg/ha) 2.30(1,15) 0.14(1,15) 0.04(1,15)
Phosphorus (%) 0.11(1,13) 1.89(1,13) 1.81(1,13)
Carbon (%) 1.01(1,15) 0.41(1,15) 0.22(1,15)
Nitrogen (%) 0.07(1,15) 0.63(1,15) 0.65(1,15)
C:N 6.16(1,15)* 0.44(1,15) 0.32(1,15)
C:P 0.31(1,13) 0.03(1,13) 0.00(1,13)
N:P 0.18(1,13) 0.15(1,13) 0.04(1,13)
Cover 0.98(1,16) 5.73(1,16)* 5.48(1,16)*
Frequency 0.00(1,16) 0.60(1,16) 1.01(1,16)
Table 4-6. Tests of fixed effects for Symplocarpus foetidus. Values are the F values and the numbers
in parentheses are the numerator and denominator degrees of freedom, respectively. N is the
nitrogen treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus
interaction effect. An asterisks (*) indicates significance at the 0.05 level.
96
Impatiens capensis
Fre
qu
en
cy
(%
)
0
5
10
15
20
25
Microstegium vimineum
Fre
qu
en
cy
(%
)
0
20
40
60
80
Polygonum punctatum
Fre
qu
en
cy
(%
)
0
5
10
15
20
(a) (b)
(c)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
(d)
Symplocarpus foetidus
Fre
qu
en
cy
(%
)
0
20
40
60
80
d)
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-7. Nitrogen and phosphorus treatment effects on the frequency of (a) Impatiens capensis,
(b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus. There were
no significant treatment effects at the 0.05 level.   ? = no nutrient added; + = nutrient added.
97
Impatiens capensis
Ph
osp
ho
rus
(%
)
0.0
0.1
0.2
0.3
0.4
0.5
Polygonum punctatum
Ph
osp
ho
rus
(%
)
0.0
0.1
0.2
0.3
0.4
0.5
Symplocarpus foetidus
Ph
osp
ho
rus
(%
)
0.0
0.1
0.2
0.3
0.4
0.5
*
Microstegium vimineum
Nitrogen Treatment
None Added
Ph
osp
ho
rus
(%
)
0.0
0.1
0.2
0.3
0.4
0.5
P0
P1
*
(a) (b)
(c) (d)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-8. Nitrogen and phosphorus treatment effects on the phosphorus concentration in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus
foetidus. Values with asterisks (*) between them indicates a significant treatment effect for that
nutrient at the 0.05 level. The nitrogen*phosphorus interaction is shown for Microstegium vimineum
since it was found to be significant (P=0.0126). The addition of phosphorus significantly increased
the concentration of phosphorus in Polygonum punctatum.   ? = no nutrient added; + = nutrient
added.
98
Impatiens capensis
Ni
tro
ge
n(
%)
0
1
2
3
4
5
Microstegium vimineum
Ni
tro
ge
n(
%)
0
1
2
3
4
5
Polygonum punctatum
Ni
tro
ge
n(
%)
0
1
2
3
4
5
Symplocarpus foetidus
Ni
tro
ge
n(
%)
0
1
2
3
4
5
(a) (b)
(c) (d)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-9. Nitrogen and phosphorus treatment effects on the nitrogen concentration in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus
foetidus. There were no significant treatment effects at the 0.05 level.   ? = no nutrient added; + =
nutrient added.
99
no significant changes in carbon concentrations caused by the addition of nitrogen or
phosphorus for any of the species studied (Figure 4-10a-d; Tables 4-3, 4-4, 4-5, and 4-6).
The N:P ratios increased for all species when nitrogen was added, but was
significantly altered only for Microstegium vimineum and Polygonum punctatum (Figure
4-11a-d; Tables 4-3, 4-4, 4-5, and 4-6). On the other hand, the N:P ratios decreased for
all species when phosphorus was added, but was significant in only the Microstegium
vimineum and Polygonum punctatum biomass (Figure 4-11a-d; Tables 4-3, 4-4, 4-5, and
4-6). There were no significant trends due to the addition of nitrogen or phosphorus in
the C:N ratios (Figure 4-12a-d; Tables 4-3, 4-4, 4-5, and 4-6). However, the C:N ratios
of Impatiens capensis, Polygonum punctatum, and Symplocarpus foetidus did decrease
somewhat when nitrogen was added (Tables 4-3, 4-5, and 4-6). The C:P ratio of all
species except Microstegium vimineum was unaffected when nitrogen was added (Figure
4-13; Tables 4-3, 4-4, 4-5, and 4-6). When nitrogen was added to Microstegium
vimineum the C:P ratio significantly increased (Figure 4-13b; Table 4-4). The C:P ratio
of this species also significantly decreased when phosphorus was added (Figure 4-13b;
Table 4-4). Although the other species did not have their C:P ratios significantly reduced
by the addition of phosphorus, their ratios did decrease somewhat (Figure 4-13a,c,d;
Tables 4-3, 4-5, and 4-6).
The biomass of Impatiens capensis, Microstegium vimineum, and Symplocarpus
foetidus were not affected by the addition of nitrogen or phosphorus (Figure 4-14a,b,d;
Tables 4-3, 4-5, and 4-6). However, the interaction of nitrogen and phosphorus did
significantly affect the biomass of Polygonum punctatum (Figure 4-14c; Table 4-5).
Since it is a rank-order interaction, meaning that the plot of the two treatment lines
100
Impatiens capensis
Ca
rbo
n(
%)
0
10
20
30
40
50
Microstegium vimineum
Ca
rbo
n(
%)
0
10
20
30
40
50
Polygonum punctatum
Ca
rbo
n(
%)
0
10
20
30
40
50
Symplocarpus foetidus
Ca
rbo
n(
%)
0
10
20
30
40
50
(a) (b)
(c) (d)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-10. Nitrogen and phosphorus treatment effects on the carbon concentration in (a)
Impatiens capensis, (b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus
foetidus.  There were no significant treatment effects at the 0.05 level.   ? = no nutrient added; + =
nutrient added.
101
Impatiens capensis
N:
P
0
5
10
15
20
Microstegium vimineum
N:
P
0
5
10
15
20
Polygonum punctatum
N:
P
0
5
10
15
20
Symplocarpus foetidus
N:
P
0
5
10
15
20
(a) (b)
(c) (d)
* *
* *
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-11. Nitrogen and phosphorus treatment effects on the N:P ratios in (a) Impatiens capensis,
(b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus. An asterisks
(*) indicates a significant treatment effect for that nutrient at the 0.05 level. The horizontal lines in
the N:P graph represent the cutoff values for nutrient limitation suggested by Koerselman and
Mueleman (1996). Values above 16 represent phosphorus limitation while those below 14 represent
nitrogen limitation. Values between 14 and 16 represent co-limitation by both nitrogen and
phosphorus. Significant treatment effects due to nitrogen and phosphorus were found for
Microstegium vimineum and Polygonum punctatum.   ? = no nutrient added; + = nutrient added. 
102
Impatiens capensis
C:
N
0
5
10
15
20
Microstegium vimineum
C:
N
0
5
10
15
20
Polygonum punctatum
C:
N
0
5
10
15
20
Symplocarpus foetidus
C:
N
0
5
10
15
20
(a) (b)
(c) (d)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-12. Nitrogen and phosphorus treatment effects on the C:N ratios in (a) Impatiens capensis,
(b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus. There were
no significant treatment effects at the 0.05 level.  ? = no nutrient added; + = nutrient added.
103
Impatiens capensis
C:
P
0
50
100
150
200
Microstegium vimineum
C:
P
0
50
100
150
200
Polygonum punctatum
C:
P
0
50
100
150
200
Symplocarpus foetidus
C:
P
0
50
100
150
200
(a) (b)
(c) (d)
* *
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-13. Nitrogen and phosphorus treatment effects on the C:P ratios in (a) Impatiens capensis,
(b) Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus. An asterisks
(*) indicates a significant treatment effect for that nutrient at the 0.05 level. Signficant treatment
effects at the 0.05 level were found for Microstegium vimineum.   ? = no nutrient added; + = nutrient
added.
104
Impatiens capensis
Bi
om
ass
(k
g/h
a)
0
10
20
30
40
50
Microstegium vimineum
Bi
om
ass
(k
g/h
a)
0
20
40
60
80
100
Symplocarpus foetidus
Bi
om
ass
(k
g/h
a)
0
500
1000
1500
2000
Polygonum punctatum
Nitrogen Treatment
None Added
Bi
om
ass
(k
g/h
a)
10
20
30
40
50
60
70
80
90
P0
P1
*
(a) (b)
(c) (d)
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Nitrogen
? + +?
Phosphorus
Treatment
Figure 4-14. Nitrogen and phosphorus treatment effects on the biomass of (a) Impatiens capensis, (b)
Microstegium vimineum, (c) Polygonum punctatum, and (d) Symplocarpus foetidus. Values with
asterisks(*) above them indicate a significant treatment effect for that nutrient at the 0.05 level.
Main effects are shown for Impatiens capensis, Microstegium vimineum, and Symplocarpus foetidus
since the nitrogen*phosphorus interactions were insignificant. The nitrogen*phosphorus interaction
is shown for Polygonum punctatum since it was found to be signficant.   ? = no nutrient added; + =
nutrient added.
105
actually intersect, no broad statements about main effects can be made.
DISCUSSION
Differences in the N:P ratio and phosphorus NUE when considering all species
combined may be attributed to significantly elevated concentrations of phosphorus in the
biomass. The plants may have been experiencing luxury uptake of this nutrient. Since
no change in biomass accompanied the increase in phosphorus uptake (indeed
phosphorus application slightly decreased biomass, phosphorus is probably not the
limiting nutrient limiting growth in this ecosystem. Nitrogen additions, however, slightly
increased biomass suggesting that the plants may be limited by nitrogen.
Phosphorus concentrations were significantly increased by phosphorus
fertilization, but neither the addition of nitrogen nor that of phosphorus created a
significant change in nitrogen or carbon concentrations. However, the nitrogen was
slightly higher when nitrogen was added. I had anticipated the concentration of nitrogen
to increase with the addition of nitrogen and the concentration of phosphorus to increase
with the addition of phosphorus. An increase in plant uptake relative to the supply of
nutrients has been documented in other experiments as well (Hubbard et al. 1999).
I did not find significant changes in biomass due to the addition of nitrogen or
phosphorus. Most fertilization experiments show that biomass is enhanced by the
addition of the limiting nutrient (Schippers and Olff 2000; G?sewell and Verhoeven
2003).  However, only 45 out of 121 plants studied by G?sewell, Koerselman, and
Verhoeven (2003) increased in biomass when nitrogen or phosphorus was added.
However, I only measured the aboveground biomass. Perhaps the belowground biomass
rather than the aboveground was increased so that the plant could better attain the
106
nutrients. Change in biomass partitioning due to fertilization have been documented in
previous studies (Claridge and Franklin 2002; Schippers and Olff 2000). Furthermore,
Aerts (1999) found that plants often either allocate more biomass to their roots or
increase their root length per unit root mass to overcome nutrient stresses.
There were no differences in percent cover or frequency due to nitrogen or
phosphorus additions. Perhaps if I had fertilized the plants several times a difference in
these two parameters would have been found. Also, I may have seen some differences if
I had measured cover and frequency prior to fertilization and again after.
Nutrient loads revealed no significant changes due to nitrogen or phosphorus
enrichment. This result makes sense when one considers the fact that no differences in
biomass were detected.
There was no significant difference in nitrogen NUE when nitrogen or
phosphorus was added. The phosphorus NUE was also unaffected by the addition of
nitrogen. Yet it was significantly decreased when phosphorus was added. The
phosphorus concentrations increased but did not produce a corresponding increase in
biomass.
Moreover, neither the C:N ratio nor the C:P ratio was significantly affected when
nitrogen or phosphorus was applied. In addition, the N:P ratio was not affected by the
addition of nitrogen but it was significantly decreased when phosphorus was added since
the phosphorus concentrations were elevated due to phosphorus fertilization. Once again
this seems to suggest that overall the community is influenced more by phosphorus than
by nitrogen.
107
Examining the results by species revealed some significant differences due to
fertilization. Phosphorus did cause the phosphorus concentrations to significantly
increase in Microstegium vimineum and Polygonum punctatum. Some suggest that if
phosphorus is limited, then the phosphorus concentrations will increase in response to the
added phosphorus (Tessier and Raynal 2003). Based on this assumption, I would
conclude that phosphorus may actually be the limiting nutrient for these two species.
However, the very definition of a limiting nutrient is that the addition of a nutrient or
combination of nutrients limits plant growth. Thus, to simply say that phosphorus is
limiting because it increased the uptake of the nutrient is not necessarily true especially
since luxury uptake could be occurring.
Microstegium vimineum did slightly increase in biomass due to both nitrogen and
phosphorus additions but since neither change is significant, it is difficult to determine
which nutrient is limiting growth. Claridge and Franklin (2002) fertilized this same plant
with all purpose (N-P-K 20-20-20) fertilizer and found that the biomass increased
significantly with the addition of the fertilizer. The biomass of Microstegium vimineum
increased slightly when I added nitrogen and phosphorus suggesting that this species may
be limited by both nutrients. However, the N:P ratios were below 14, the cutoff value
representing nitrogen limitation suggested by Koerselman and Meuleman (1996). Thus, I
believe that this species is limited by nitrogen. This is especially important for
Microstegium vimineum since this plant is an introduced species. The interaction of
nitrogen and phosphorus was found to cause a significant difference in the biomass of
Polygonum punctatum. However, since this interaction is rank-order the relationship of
nitrogen and phosphorus to the biomass change cannot be distinguished. An increase in
108
plant biomass is often observed when the limiting nutrient is added. However, G?sewell,
Koerselman, and Verhoeven (2003) found results that out of 121 plants studied, only 45
had an increase in biomass due to the addition of nitrogen or phosphorus. There were no
differences in nitrogen or carbon concentrations in any of the species due to the addition
of nitrogen or phosphorus.
Moreover element ratios were inspected so that species responses due to nitrogen
or phosphorus could be further extrapolated. The N:P ratios of Microstegium vimineum
and Polygonum punctatum were both significantly increased by the application of
nitrogen. The N:P ratios of both these species were also decreased significantly when
phosphorus was applied. The N:P ratios of the other species were not significantly
influenced by nitrogen or phosphorus. C:N ratios are often used to determine if nitrogen
is limited in quantity or readily available in ecosystems. It was anticipated that the C:N
ratios of all species would decrease in the plots receiving nitrogen if this was the limiting
nutrient. All of the species except Microstegium vimineum decreased somewhat when
nitrogen was added. Thus, these results suggest that Impatiens capensis, Polygonum
punctatum, and Symplocarpus foetidus were taking up nitrogen thereby decreasing the
C:N ratio. Investigation of C:P ratios revealed that the addition of nitrogen was not
expected to produce a change in the C:P ratios. However, the C:P ratio in Microstegium
vimineum biomass was significantly increased when nitrogen was added. This species is
extremely plastic (i.e. can adapt easily to changes in environmental conditions) (Claridge
and Franklin 2002). Perhaps it produced more carbon when nitrogen was added due to an
increase in photosynthesis causing an increase in the C:P ratio. Indeed, the carbon
concentration in this species was slightly increased by the addition of nitrogen and this is
109
probably causing the C:P ratio to change. The C:P ratio in Microstegium vimineum was
also significantly decreased by the addition of phosphorus since the phosphorus
concentration increased significantly.
No differences were observed in percent cover or frequency by species when
nitrogen or phosphorus was applied. Joyce (2001) studied the effects of nitrogen
fertilization on the vegetation in a flood-meadow. He found a wide variety of results.
Fertilization caused some grass species to increase in cover and some to be unaffected at
all. However, the majority of species actually decreased in cover. Perhaps the plants
respond slowly to added nutrients. I may have been able to detect significant differences
if I continued to fertilize the plants for several growing seasons.
CONCLUSIONS
When looking at the effect of fertilization on the entire plant community, only
phosphorus caused significant changes. The N:P ratio and phosphorus NUE of the entire
plant community were significantly decreased when phosphorus was applied due to the
significantly higher concentrations of phosphorus taken up by plant. Microstegium
vimineum and Polygonum punctatum were the only species that experienced significant
changes due to fertilization. Moreover, these shifts involved nitrogen and phosphorus
only (i.e. P concentrations and N:P and C:P ratios). Thus, the addition of nutrients may
affect the normal nutrient cycling with respect to these two species in particular.
More long-term fertilization experiments must be conducted to determine the
whether forested riparian wetlands are limited by nitrogen, phosphorus, or a combination
of the two nutrients. The effects of chronic nutrient additions on plant communities
should be further investigated so that better management decisions can be made. By
110
studying these systems more thoroughly, others may be able to better predict the
consequences of agricultural development using computer modeling techniques.
111
Chapter 5: The Effects of Nutrient Enrichment on
Symplocarpus foetidus
INTRODUCTION
Nonpoint source pollution caused by excess nutrients is a continuing problem in
the United States (Carpenter et al. 1998). The level of nutrients entering waterways is
continuing to increase as agricultural and suburban areas encroach upon natural
ecosystems. Controlling nitrogen and phosphorus loading is particularly important since
these two nutrients often lead to eutrophication.
The vegetation in these areas is especially important since plants actively take up
nutrients. Silvan et al. (2004) found that the vegetation in a constructed wetland buffer
was the primary factor mitigating nitrogen and, to a lesser degree, phosphorus entering
from nearby forestry production. Studying the direct effects of nitrogen and phosphorus
on wetland species found in natural riparian wetland buffers is necessary so that the role
of vegetation in nutrient retention can be determined and may ultimately lead to better
management practices.
Symplocarpus foetidus (L.) Salisb. ex Nutt (skunk cabbage) is an herbaceous
species that has unique characteristics. Its respiration produces heat up to a temperature
of 72?F in the winter when air temperatures are low and is one of the first plants to
emerge in the spring (Eastman 1995). Its large leaves make it easily identifiable
taxonomically and the high surface area of the leaves can potentially allow the plant to
harvest a large amount of sunlight for photosynthesis. Moreover, it has thick roots that
extend deep into the soil profile (approximately equal to the height of the plant- usually
about 0.5 m) making them difficult to remove from the soil (personal observation).
112
Indeed, the roots contract slightly acting to pull the plant deeper in the soil by several
millimeters every year (Eastman 1995). These roots are even thought to persist for
centuries (Eastman 1995). Because of their characteristic roots and large leaves, I
thought that this plant had a high potential for attenuating large amounts of nutrients.
Despite these unique traits, there is little, if any, literature investigating the nutrient
limitation and nutrient uptake of this species. I conducted an experiment to examine the
role of Symplocarpus foetidus in nutrient uptake by fertilizing these plants in situ in a
forested riparian wetland not receiving direct additions of nutrients.
I anticipated that the addition of nitrogen and phosphorus would increase the
nutrient concentrations in the leaves since more nitrogen would be available for uptake. I
also thought biomass would increase when the limited nutrient(s) was added since this is
a typical response shown by fertilized plants. It was also thought that the leaf area, plant
width, and plant height would increase since an increase in biomass would probably
mean an increase in the size of the leaf and plant in general. Total nitrogen and carbon
and inorganic nitrogen forms (ammonium and nitrate) were also measured in the soil.
Concentrations o all forms of nitrogen were anticipated to increase due to the application
of nitrogen.
METHODS
In June of 2003, forty Symplocarpus foetidus plants were selected to be fertilized
in a forested riparian wetland that was not exposed to high nutrient concentrations
(reference site). Selection of the plants was based on similarities in general size, the
number of leaves, and height. Ten plants were randomly assigned to each of the four
113
treatments, which included (1) the addition of nitrogen, (2) the addition of phosphorus,
(3) addition of both nitrogen and phosphorus, and (4) no additions.
Nitrogen was added in the form of urea fertilizer (N-P-K (46-0-0 ratio))and
phosphorus was added in the form of triple superphosphate (N-P-K (0-46-0)). One
hundred and fifty grams of the assigned treatment was applied twice to account for any
possible delays in the plants response to the additional nutrients. The first application
occurred on June 19, 2003 (the middle to end of the 2003 leaf growth season for
Symplocarpus foetidus) and the second was on April 6, 2004 (the beginning of the 2004
leaf growth season for Symplocarpus foetidus). The area chosen for this experiment was
frequently flooded. The fertilizer was placed below the soil surface to a depth of 20 cm
using a soil corer to prevent the loss of fertilizer during these flooding events. Holes
were also dug under plants not receiving fertilizer to account for any possible influences.
The holes were then sealed with soil.
I examined several parameters to determine if fertilization affected the general
growth of the plants. The maximum plant height and width and the length of the
youngest leaf, which was tagged on the first measurement date, were measured.
Measurements of tagged leaves were taken on May 12, 2004 and again after two weeks.
Leaf area indices (LAI) were also measured 20 cm from the base of each plant and 5 cm
above the ground using a LI-COR 1 m quantum sensor with an LI-250 light meter (LI-
COR, Lincoln, Nebraska). LAI is a dimensionless number representing the area of
foliage per unit area of ground surface and is therefore related to plant biomass.
The plant leaves were harvested at the end of the 2004 growing season prior to
senescence and placed in a 60?C oven until constant weight was achieved.  The samples
114
were then weighed and homogenized using a Wiley Mill. Total phosphorus was
extracted using a modified Kjeldahl digestion method (5 mL of sulfuric acid was added to
each sample instead of the 3.5 mL called for in the Kjeldahl digestion protocol) and
analyzed on a Flow Injection Analyzer (Lachat QuikChem 8000, Milwaukee,
Wisconsin). Total nitrogen and carbon were also determined via dry combustion using a
LECO Tru-Spec CN Analyzer (St. Louis, Missouri).
Soil samples were also collected to ensure that the fertilizer had remained at the
site of application and available for plant use. Two samples were obtained underneath
each plant to a depth of 20 cm using a handheld soil extractor. The two samples were
then combined to make one sample per plant, air dried, and homogenized using a mortar
and pestle. A LECO Tru-Spec CN Analyzer (St. Louis, Missouri) was used to ascertain
the total nitrogen and carbon in the samples. Inorganic nitrogen in the form of
ammonium and nitrate was determined by first extracting the nitrogen using a 2 M KCl
solution (Mulvaney 1996) and then analyzing it on a Flow Injection Analyzer (Lachat
QuikChem 8000, Milwalkee, Wisconsin). C:N ratios were computed using the total
nitrogen and carbon data on a percentage basis. Soil phosphorus concentrations were not
determined mainly because I did not have the available time or the necessary resources to
complete the analyses. Phosphate determinations via the dilute acid extraction procedure
(also known as North Carolina or Mehlich-1 P test) require digesting the samples with a
strong acid in a specialized hood (Kuo 1996). The specialized hood required to safely
complete the analyses was unavailable for use until a later time because it was being used
by other researchers. Phosphorus is transported mainly through surface processes by
adhering to soil particles and traveling in surface runoff. Since I placed the fertilizer
115
below the soil surface, I felt that the loss due to possible flooding would be minimal.
RESULTS
The application of nitrogen produced a significant increase in both the leaf area
index (Figure 5-1; Table 5-1) and concentrations of nitrogen in the leaves (Figure 5-2a;
Table 5-1). The phosphorus treatment had a significantly positive effect on the
phosphorus concentrations in the leaves (Figure 5-2b; Table 5-1). Nevertheless, leaf
carbon concentrations were unaffected by either nitrogen or phosphorus (Figure 5-2c).
In addition, there were some differences found for element ratios when treatments
were applied. The N:P ratios were significantly increased when nitrogen was added
(Figure 5-3a; Table 5-1). However, both the N:P (Figure 5-3a; Table 5-1) and C:P ratios
(Figure 5-3b; Table 5-1) were significantly decreased when phosphorus was added. The
C:N ratios (Figure 5-3c; Table 5-1), on the other hand, were not altered by the nutrient
treatments.
The interaction of nitrogen and phosphorus produced a significant effect on leaf
width (Figure 5-4; Table 5-2). The application of nitrogen and phosphorus together
caused the plant width to be significantly greater than when nitrogen alone was applied.
Surprisingly, fertilization did not affect plant height (Figure 5-5a; Table 5-2), the length
of the youngest leaf (Figure 5-5b; Table 5-2), or biomass (Figure 5-6; Table 5-2).
The total nitrogen in the soil was significantly increased when nitrogen was
applied (Figure 5-7a; Table 5-3). The application of nitrogen also significantly increased
both plant ammonium (Figure 5-7b; Table 5-3) and nitrate (Figure 5-7c; Table 5-3) in the
116
LA
I
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
*
? + ? +
Treatment
Nitrogen Phosphorus
Figure 5-1. Nitrogen and phosphorus treatment effects on the leaf area indices (LAI) of
Symplocarpus foetidus. Values with asterisks (*) above them indicate a significant treatment effect
for that nutrient at the 0.05 level. The LAI increased significantly when nitrogen was applied, but
was unaffected by the application of phosphorus.  ? = no nutrient added; + = nutrient added.
117
Parameter N P N*P
LAI 6.10(1,36)* 0.84(1,36) 0.14(1,36)
Biomass 0.01(1,36) 3.25(1,36) 0.00(1,36)
Nitrogen 4.60(1,36)* 0.56(1,36) 0.08(1,36)
Phosphorus 1.74(1,36) 8.36(1,36)** 0.55(1,36)
Carbon 1.14(1,35) 0.02(1,35) 1.34(1,35)
N:P 9.17(1,36)** 21.49(1,36)*** 0.02(1,36)
C:P 2.17(1,36) 10.26(1,36)** 0.06(1,36)
C:N 3.36(1,36) 0.95(1,36) 0.02(1,36)
Table 5-1. Tests of fixed effects for LAI, biomass, nitrogen, phosphorus, and carbon concentrations,
and N:P, C:P, and C:N ratios in Symplocarpus foetidus. Values are the F values and the numbers in
parentheses are the numerator and denominator degrees of freedom, respectively. N is the nitrogen
treatment effect; P is the phosphorus treatment effect; N*P is the nitrogen*phosphorus interaction
effect. An asterisks (*) indicates significance at the following levels: * = <0.05; ** = < 0.01; *** = <
0.0001.
118
Ni
tro
ge
n(
%)
0
1
2
3
4
Ph
osp
ho
rus
(%
)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Ca
rbo
n(
%)
0
10
20
30
40
50
*
*
? + ? +
Treatment
Nitrogen Phosphorus
(a)
(b)
(c)
Figure 5-2. Nitrogen and phosphorus treatment effects on the concentrations of (a) nitrogen, (b)
phosphorus, and (c) carbon in the leaves of Symplocarpus foetidus. Values with asterisks (*) between
them indicate a significant treatment effect for that nutrient at the 0.05 level. The concentration of
nitrogen increased significantly when nitrogen was applied and the concentration of phosphorus
increased significantly when phosphorus was added. Carbon concentrations were unaffected by
fertilization.  ? = no nutrient added; + = nutrient added.
119
C:
N
0
2
4
6
8
10
12
14
C:
P
0
50
100
150
200
250
N:
P
0
2
4
6
8
10
12
14
16
18
*
* *
? + ? +
Treatment
Nitrogen Phosphorus
(a)
(b)
(c)
Figure 5-3. Nitrogen and phosphorus treatment effects on the (a) N:P, (b) C:P, and (c) C:N element
ratios. Values with asterisks (*) above them indicate a significant treatment effect for that nutrient
at the 0.05 level. The horizontal lines in the N:P graph represent the cutoff values for nutrient
limitation suggested by Koerselman and Mueleman (1996). Values above 16 represent phosphorus
limitation while those below 14 represent nitrogen limitation. Values between 14 and 16 represent
co-limitation by both nitrogen and phosphorus. The application of nitrogen had no effect on the C:N
or C:P ratios, but significantly increased the N:P ratio. The application of phosphorus had not effect
on the C:N ratio, but significantly decreased the C:P and N:P ratios.  ? = no nutrient added; + =
nutrient added.
120
Nitrogen Treatment
None Added
W
idt
h(
cm
)
80
85
90
95
100
105
P1
P0
*
Figure 5-4. Nitrogen and phosphorus treatment effects on plant width. The interaction of
nitrogen*phosphorus was found to be significant using the tests of fixed effects and so is presented
here. Values with asterisks (*) above them indicate a significant treatment effect for that nutrient at
the 0.05 level. The addition of both nitrogen and phosphorus caused the plant width to be
significantly higher than when just nitrogen was added.  ? = no nutrient added; + = nutrient added.
Parameter N P T N*P N*T P*T N*P*T
Length 0.49(1,63) 0.64(1,63) 10.05(1,63)** 0.58(1,63) 0.10(1,63) 0.01(1,63) 0.179(1,63)
Width 0.19(1,63) 4.53(1,63)* 4.28(1,63)* 6.86(1,63)* 0.03(1,63) 0.00(1,63) 0.26(1,63)
Height 0.54(1,63) 0.06(1,63) 1.04(1,63) 0.01(1,63) 0.00(1,63) 0.04(1,63) 0.00(1,63)
Table 5-2. Tests of fixed effects results for length, width, and height of Symplocarpus foetids. Values are F values and the numbers in parentheses are
the numerator and denominator degrees of freedom, respectively. N is the nitrogen treatment effect; P is the phosphorus treatment effect; T is the time
effect; N*P is the nitrogen*phosphorus interaction effect; N*T is the nitrogen*time interaction; P*T is the phosphorus*time interaction; N*P*T is the
nitrogen*phosphorus*time effect. An asterisks (*) indicates significance at the following levels: * = <0.05.
122
Le
ng
th
(cm
)
0
10
20
30
40
50
60
? + ? +
Nitrogen Phosphorus
He
igh
t(c
m)
0
20
40
60
80
(a)
(b)
Figure 5-5. Nitrogen and phosphorus effects on (a) plant height and (b) the length of the youngest
leaf.  There were no significant treatment effects at the 0.05 level.  ? = no nutrient added; + =
nutrient added.
123
Plot 1
Bi
om
ass
(g
)
0
1
2
3
4
5
? + ? +
Treatment
Nitrogen Phosphorus
Figure 5-6. Nitrogen and phosphorus effects on plant biomass. There were no significant treatment
effects at the 0.05 level.  ? = no nutrient added; + = nutrient added.
124
Ni
tro
ge
n(
%)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
*
Am
mo
niu
m
(pp
m)
0
20
40
60
80
100
*
Ni
tra
te
(pp
m)
0
5
10
15
20
25
*
? + ? +
Treatment
Nitrogen Phosphorus
(a)
(b)
(c)
Figure 5-7. Nitrogen and phosphorus treatment effects on (a) total nitrogen, (b) ammonium, and (c)
nitrate concentrations in soil. Values with asterisks (*) between them indicate a significant
treatment effect for that nutrient at the 0.05 level. The addition of nitrogen significantly increased
the concentrations of all the nitrogen species.  ? = no nutrient added; + = nutrient added.
125
Parameter N P N*P
Total Nitrogen 5.73(1,36)* 0.59(1,36) 0.27(1,36)
Nitrate 25.57(1,36)** 1.26(1,36) 1.02(1,36)
Ammonium 26.53(1,36)** 3.54(1,36) 2.73(1,36)
Total Carbon 6.24(1,36)* 0.02(1,36) 0.02(1,36)
C:N 41.40(1,36)** 0.73(1,36) 0.04(1,36)
Table 5-3. Tests of fixed effects results for soil. Values are F values and the numbers in parentheses
are the numerator and denominator degrees of freedom, respectively. N is the nitrogen treatment
effect; P is the phosphorus treatment effect; T is the time effect; N*P is the nitrogen*phosphorus
interaction effect. An asterisks (*) indicates significance at the following levels: * = <0.05; ** = <
0.0001.
126
soil. However, carbon concentrations in the soil actually significantly decreased with the
addition of nitrogen (Figure 5-8; Table 5-3).
Figure 5-8. Nitrogen and phosphorus treatment effects on the carbon concentrations in soil. Values
with asterisks (*) between them indicate a significant treatment effect for that nutrient at the 0.05
level. The addition of nitrogen significantly decreased carbon concentrations.  ? = no nutrient
added; + = nutrient added.
Because of these changes in nutrient concentrations associated with added nitrogen, the
C:N ratio of the soil also significantly decreased due to the nitrogen treatment (Figure 5-
9; Table 5-3).
DISCUSSION
The leaf area of the plants increased due to the addition of nitrogen. This result
was of great interest since it is related to plant biomass. I also thought that plant height
and width along with the length of the youngest leaf would increase in response to the
added nutrients. Hubbard et al. (1999) studied the influence of swine lagoon effluent on
six wetland plant species. They found that some of the species, especially broad-leaved
plants, that were exposed to the swine effluent increased in leaf area, plant height, and
Ca
rbo
n(
%)
0
1
2
3
4
5
? + ? +
Treatment
Nitrogen Phosphorus
*
127
C:
N
0
2
4
6
8
10
12
14
16
18
? + ? +
Treatment
Nitrogen Phosphorus
*
Figure 5-9. Nitrogen and phosphorus treatment effects on the C:N ratios in soil. Values with
asterisks (*) above them indicate a significant treatment effect for that nutrient at the 0.05 level. The
addition of nitrogen significantly decreased C:N ratios.  ? = no nutrient added; + = nutrient added.
128
growth index, which is a calculation involving plant height and width (Hubbard et al.
1999). However, I may not have found a change in plant height and width or the length
of the youngest leaf because Symplocarpus foetidus may respond slowly to the additions
of nutrients. In addition, its large roots and corm may store large stocks of nutrients.
This increase in leaf area has important implications since the plants are effectively
increasing the area exposed to sunlight and able to photosynthesize.
In addition, the nitrogen and phosphorus in the leaves significantly increased in
response to added nitrogen and phosphorus, respectively. This increase in plant uptake
relative to the supply of nutrients has been documented in other experiments as well
(Hubbard et al. 1999). I expected to find a difference in carbon concentrations due to
nitrogen and/or phosphorus additions. I thought that whatever nutrient(s) was limiting
plant growth would cause a change in plant biomass, which would result in an increase in
carbon production and hence carbon concentrations. However, maybe the absolute
carbon amounts increased instead of the concentrations.
Both the N:P and C:P ratios were anticipated to decrease when phosphorus was
added. It was believed that higher phosphorus availability would cause uptake to
increase and ultimately lead to higher concentrations in the leaves. These higher
phosphorus concentrations then caused both the N:P and C:P ratios to decrease. N:P
ratios were also influenced by nitrogen. Once nitrogen was added, the concentrations of
nitrogen in the leaves increased, yielding larger N:P ratios. Koerselman and Meuleman
(1996) proposed that N:P ratios greater than 16 indicate phosphorus limitation, those
below 14 represent nitrogen limitation, and those values in between reflect co-limitation
by nitrogen and phosphorus. The N:P ratio increased from 12.5 to 14.6 when nitrogen
129
was added. This result suggests that nitrogen addition resulted in a shift from nitrogen
limitation to co-limitation by nitrogen and phosphorus. Similarly, when phosphorus was
added, the plants shifted from 15.1 to 12.0 reflecting since they had higher concentrations
of phosphorus in their leaves. No differences in C:N ratios were found but the addition
of nitrogen did decrease it slightly.
Although I did not find any differences in biomass due to the treatments, other
studies have concluded the same results.  Only 45 out of 121 plants studied by G?sewell,
Koerselman, and Verhoeven (2003) increased in biomass when nitrogen or phosphorus
was added. Since Symplocarpus foetidus have high belowground biomass, perhaps this
fraction of the biomass rather than the aboveground was increased so that the plant could
better attain the nutrients. Changes in biomass partitioning due to fertilization have been
documented in previous studies (Schippers and Olff 2000; Claridge and Franklin 2002).
The addition of nitrogen significantly increased all of the nitrogen components in
the soil (total nitrogen, ammonium, and nitrate). This expected result indicates that the
fertilizer was not washed out of the system. It has been shown that wetland soils often
provide long-term storage of nutrients (Bischoff et al. 2001). Moreover, both nitrification
and denitrification are important mechanisms of nutrient retention and removal in the soil
(Mitsch and Gosselink 1993). Unexpectedly, the addition of nitrogen also impacted the
concentrations of carbon?negatively. This result could be due to several things.
Perhaps the added nitrogen actually caused some of the carbon in the form of organic
matter to breakdown and thus was removed from the system. Also, it is known that
Symplocarpus foetidus plants are thermoregulators (Seymour and Baylock 1999; Ito et al.
2003). For a period of roughly two weeks in the winter months these plants produce heat.
130
Maybe the addition of nitrogen acted to promote the stimulation of microbes involved in
decomposition or somehow increased the length of time that the plant released heat.
Since increased temperatures can promote the breakdown of organic matter through
decomposition, perhaps it led to the removal of more organic matter. However, it could
just be due to random error. The C:N ratios were also significantly decreased by the
application of nitrogen. This change has important implications for soil microbes.
CONCLUSIONS
Increased nitrogen and phosphorus caused distinct changes in Symplocarpus
foetidus to occur. The plants responded to these excess nutrients by taking up more
nitrogen and phosphorus. By doing so their N:P and C:P ratios were also altered. The
plant shifted more towards co-limitation by nitrogen and phosphorus when nitrogen was
added but towards nitrogen limitation when phosphorus was added. This has important
implications for the cycling of nutrients in the ecosystem as a whole since this ratio can
affect soil processes and microbial communities. Determining the long-term impact of
excess nutrients on wetland ecosystems will lead to better management practices and a
greater understanding of the role of vegetation in nutrient retention within agricultural
watersheds.
131
Chapter 6: Synthesis and Conclusions
The ability of forested riparian wetlands to mitigate nutrients and chemicals has
been well documented. The role of vegetation in this buffering capacity, however, has
been long overlooked as a major influencing factor. Indeed, the vegetation has long been
considered to be less important than other factors, mainly soils, since it was thought to
provide only temporary retention of nutrients through uptake. However, the impact of
plants in the processing of excess nutrients can no longer be neglected. Results from the
combination of observational studies and fertilization experiments that I conducted
suggest that excess nutrients impacted the plant nutrient concentrations and consequently
the plant component of nutrient cycling. In addition, the plant community had lower
species diversity, richness, and evenness in the agriculturally influenced wetland when
compared to the non-enriched wetland.
The N:P ratios of herbaceous vegetation at both systems were investigated to
determine which nutrient (or combination of nutrients) was limiting. Observational
studies revealed that the overall N:P ratios of both sites were less than 14 suggesting that
the wetlands were limited by nitrogen. Drought appears to influence the N:P ratios in
herbaceous vegetation. The N:P ratio was lower in 2002 (drought year) than 2003
(?normal? year) at both sites possibly indicating that during periods of drought, nitrogen 
limited sites may actually become further restricted by nitrogen. Moreover, the N:P ratio
was significantly higher at the reference site than at the agricultural site in 2002 but was
only slightly higher in 2003. The agricultural site used in my research was found to be
largely influenced by groundwater. Therefore, during a drought there would be a lower
132
supply of nitrogen entering the system and available for plant use. While
counterintuitive, this result could mean that drought may cause enriched nitrogen limited
sites to have lower N:P ratios.
The direct application of nitrogen and phosphorus to vegetation found at the
reference site provided results that may be important in plant conservation and
management. The fertilization of all plants in vegetative plots produced significant
results due to phosphorus additions when all species were examined together by
treatment. The concentrations of phosphorus in the aboveground biomass were
significantly increased while both the phosphorus NUE and the N:P ratios were
significantly decreased when phosphorus was applied. These results could be due to
luxury uptake of phosphorus. If phosphorus was limiting this plant community, one
would expect the overall plant biomass to increase when this nutrient was added.
Although not significant, aboveground biomass decreased slightly when phosphorus was
applied and increased slightly with the addition of nitrogen. Thus, this supports the
conclusion that the reference is nitrogen limited.
Symplocarpus foetidus may be a good candidate for use in created riparian
wetlands used specifically for nutrient mitigation since those found at the agricultural site
had significantly higher concentrations of nitrogen and phosphorus in 2003 than those at
the reference site. In the Symplocarpus foetidus fertilization experiment, the plant uptake
of nitrogen and phosphorus were both increased in leaf tissue with the addition of
nitrogen and phosphorus, respectively. The LAI was also significantly increased by the
addition of nitrogen indicating a change in growth pattern. In addition, the percent cover
was significantly higher when only nitrogen was applied compared to when both nitrogen
133
and phosphorus were applied. Although not significant, the frequency of occurrence was
also slightly higher in plots receiving nitrogen fertilizer. The cover and frequency
estimates that were determined in the observational study for Symplocarpus foetidus
further support these findings since both parameters were significantly higher at the
agricultural site than at the reference site.
While both percent cover and frequency may be influenced by light, nitrogen, and
water, it appears that the growth of this species is particularly influenced by water supply
and nitrogen availability. In 2003, the frequency of Symplocarpus foetidus was almost
100% in all of the plots at the agricultural site, except those in the plot furthermost
upstream. In this plot the frequency was 62.5%. This is important since this upstream
area is extremely narrow in width and consequently has a high amount of light
penetrating the canopy. If this species were limited mainly by light in this environment, I
would expect it to have the highest cover and frequency in this plot. Furthermore, the
surface runoff from the agricultural fields does not drain into this upper region nor does it
receive a lot of groundwater input when compared to some of the other plots. Since
nitrogen is transported via these two mechanisms, the nitrogen supply may be more
limited in this upper region. This suggests that nitrogen and water availability may be the
main factors limiting the growth of this species in this environment. Thus, if natural
areas containing Symplocarpus foetidus are considered for use in mitigating nitrogen, it
can be anticipated that this species would respond by increasing in growth, cover, and
possibly frequency as long as the source of water was not limited.
The influence of excess nutrients on Microstegium vimineum, a non-native
species, was of particular interest. By fertilizing plant communities I was able to
134
determine which nutrient was limiting growth and colonization of individual species
within the ecosystem. N:P ratios indicated that this species was limited by nitrogen. A
biomass response was not found, but did increase slightly with the addition of nitrogen
and phosphorus. The N:P ratios in this species? aboveground biomass that were 
determined in the observational study further supported that nitrogen was the limiting
nutrient. Therefore, Microstegium vimineum would most likely spread in habitats that
were impacted by excess nitrogen.
Additionally, overall comparisons between the study sites in 2003 revealed that
the concentrations of nitrogen, phosphorus, and carbon were significantly higher in tree
leaf litter, but not in aboveground herbaceous vegetation, at the agricultural site. It has
been documented by other researchers that the groundwater nitrate concentrations at the
agricultural site increase with depth. Since tree roots grow deeper than the herbaceous
plant roots, the trees take up groundwater that is deeper in depth and therefore has more
nutrients, particularly nitrogen. Also, the trees may be allocating these resources to their
aboveground biomass more so than their herbaceous counterparts. The herbaceous plants
in contrast may be allocating the acquired nutrients to belowground biomass in an effort
to further extract nutrients from the soil. The impact of nutrients on the belowground
biomass of plants must be studied to verify this possibility. Further studies of herbaceous
and tree litter must also be performed to determine if elevated nutrients in plant tissues
alter soil nutrient processes like decomposition and denitrification rates. For instance, the
elevated concentrations in the plant litter may actually promote soil mineralization rates
since more nitrogen is available to stimulate the microbes involved in this process.
135
Furthermore, the NUEs of nitrogen and phosphorus were both significantly higher
in the agricultural tree litter than the reference litter. This suggests that trees in nutrient-
rich environments may be less efficient at retaining nutrients from leaves than those in
non-impacted areas. High NUE could therefore be an adaptation to low nutrient
availability. Further research investigating the relationship of elevated nutrients and
NUEs is needed.
This research also indicates that trees found in nutrient enriched sites resorb
nutrients better during the latter period of senescence. The leaf litter at the agricultural
site had higher concentrations at the beginning of the 2003 fall season and significantly
decreased during the senescent period whereas those concentrations in the reference litter
remained relatively low and stable. By the end of the season the agricultural trees had
resorbed the excess nutrients in the leaves leaving their concentrations approximately
equal to those at the reference site. The excess nutrients in the agricultural litter may
therefore be altering the nutrient cycling of litter.
The ability of herbaceous aboveground biomass to take up nutrients was found to
vary by species. This result may be due to several factors including, but not limited to,
the following: (1) certain plant species may be more adept at harvesting and utilizing
nutrients; (2) species may differ in their allocation of nutrients to different parts of the
plant (belowground versus aboveground) due to annual variation (i.e. climactic changes)
or differing life histories (i.e. species primarily focused on competition versus those
focused on reproduction); and (3) plants may take up more nutrients than necessary for
plant sustenance (i.e. ?luxury uptake?).  Yearly variation in nutrient concentrations was 
also found. This reflects the importance of studying patterns of nutrient cycling over
136
several years. I suspect that the data from 2003 is more indicative of normal nutrient
cycles within the plant communities rather than the data from 2002, which was a drought
year.
Furthermore, the influence of excess nutrients on the plant community structure
within forested riparian wetlands must be determined so that these buffers may be better
managed to ensure the continued mitigation of pollutants. The agricultural wetland in
this study was typified by lower overall species diversity, richness, and evenness when
compared to the reference site. This implies that some plants within these nutrient
enriched wetlands may be replaced by species that are better able to cope with excess
nutrients.
Forested riparian wetlands provide ample opportunities for nutrient retention to
occur. By continuing to study the relationships of plants, water, and soils in these
ecosystems we will be better able to predict their responses to increased fertilization. It
is clear that the plant community structure and the plant nutrient dynamics are affected by
elevated nutrients. Finding ways to preserve and maintain the vegetation in these in these
areas is of utmost importance to maintaining the ecosystem services they provide. In
conclusion, the main findings of this research were:
?Agricultural sites may have lower species diversity, richness, and evenness due to
elevated nutrients.
?The trees found in agricultural settings may be larger due to excess nutrients.
?Fragmentation of wetlands in agricultural settings may reduce the total litter input
of these systems.
?Tree leaf litter at agricultural sites has elevated nutrient concentrations.
137
?Nutrient loading may not be higher at agricultural sites due to lower litter input
caused by fragmentation.
?It appears that both communities are nitrogen limited.
?Elevated nutrient concentrations in deep groundwater at agricultural sites may
result in higher concentrations in tree leaf litter but not in herbaceous vegetation.
?Trees in nutrient-rich environments may retain a smaller proportion of nutrients
from leaves than those in non-impacted areas (or that there is an upper limit to
resorption rate).
?Symplocarpus foetidus may be nitrogen-limited in forested riparian wetlands.
138
Appendix A
Table A-1. List of the scientific and common names, species codes, families, and wetland
indicator status for herbaceous plants studied at both sites.
(Page 1 of 3)
Scientific Name Common Name SpeciesCode Family
Wetland
Indicator
Status
Apios americana Groundnut APAM Fabaceae (Pea family) FACW
Arisaema triphyllum (L.)
Schott Jack in the pulpit ARTR Araceae (Arum family) FACW-
Boehmeria cylindrica(L.) Sw. Smallspike falsenettle BOCY Urticaceae (Nettle family) FACW+
Carex intumescens Rudge Greater bladdersedge CAIN12 Cyperaceae (Sedge family) FACW+
Carex lurida Wahlenb. Shallow sedge CALU5 Cyperaceae (Sedge family) OBL
Cinna arundinacea L. Sweet woodreed CIAR2 Poaceae (Grass family) FACW+
Claytonia virginica Virginia springbeauty CLVI Portulacaceae (Purslane family) FACU
Dioscorea villosa L. Wild yam DIVI Dioscoreaceae (Yam family) FAC+
Duchesnea indica (Andr.)
Focke Indian strawberry DUIN Rosaceae (Rose family) FACU-
Fragaria virginiana Duchesne Virginia strawberry FRVI Rosaceae (Rose family) FACU
(Continued)
139
(Page 2 of 3)
Scientific Name Common Name SpeciesCode Family
Wetland
Indicator
Status
Impatiens capensis Meerb. Jewelweed IMCA Balsaminaceae (Touch-me-notfamily) FACW
Juncus effusus L. Common rush JUEF Juncaceae (Rush Family) FACW+
Leersia oryzoides (L.) Sw. Rice cutgrass LEOR Poaceae (Grass family) OBL
Microstegium vimineum
(Trin.) A. Capus Nepalese browntop MIVI Poaceae (Grass family) FAC
Onoclea sensibilis L. Sensitive fern ONSE Dryopteridaceae (Wood fernfamily) FACW
Osmunda cinnamomea L. Cinnamon fern OSCI Osmundaceae (Royal fern family) FACW
Panicum clandestinum (L.)
Gould Deer-tongue grass PACL Poaceae (Grass family) FAC+
Panicum dichotomum L. Fall panicgrass PADI Poaceae (Grass family) FACW-
Parthenocissus quinquefolia
(L.) Virginia creeper PAQU Vitaceae (Grape family) FACU
Panicum virgatum L, Switchgrass PAVI2 Poaceae (Grass family) FAC
Peltandra virginica (L.) Schott Green arrow arum PEVI Araceae (Arum family) OBL
Pilea pumila (L.) Gray Canadian clearweed PIPU Urticaceae (Nettle family) FACW
Polygonum arifolium L. Halberdleaftearthumb POAR Polygonaceae (Buckwheat family) OBL
Polygonum persicaria L. Spotted ladysthumb POPE Polygonaceae (Buckwheat family) FACW
Polygonum punctatum Ell. Dotted smartweed POPU Polygonaceae (Buckwheat family) OBL
(Continued)
140
(Page 3 of 3)
Scientific Name Common Name SpeciesCode Family
Wetland
Indicator
Status
Polygonum sagittatum L. Arrowleaftearthumb POSA Polygonaceae (Buckwheat family) OBL
Sagittaria latifolia Willd. Broadleafarrowhead SALA2 Alismataceae (Water-plantainfamily ) OBL
Scutellaria lateriflora L. Blue skullcap SCLA Lamiaceae (Mint family) FACW+
Schoenoplectus
tabernaemontani (K.C. Gmel.)
Palla
Softstem bulrush SCTA2 Cyperaceae (Sedge family) OBL
Smilax rotundifolia L. Roundleafgreenbrier SMRO Smilacaceae (Catbrier family) FAC
Symplocarpus foetidus (L.)
Salisb. ex Nutt. Skunk cabbage SYFO Araceae (Arum family) OBL
141
Table A-2. List of the scientific and common names, species codes, families, and wetland
indicator status for woody plants studied at both sites.
Scientific Name Common Name
Species
Code Family
Wetland
Indicator
Status
Acer rubrum L. Red maple ACRU Aceraceae (Maple family) FAC
Betula nigra L. River birch BENI Betulaceae (Birch family) FACW
Ilex opaca Ait. American holly ILOP Aquifoliaceae (Holly family) FACU+
Ilex verticillata (L.) Gray
Common
winterberry ILVE Aquifoliaceae (Holly family) FACW+
Liquidambar styraciflua L. Sweetgum LIST Hamamelidaceae (Witch-hazelfamily) FAC
Lyonia ligustrina (L.) DC. Maleberry LYLI Ericaceae (Heath family) FACW
Nyssa sylvatica Marsh Blackgum NYSY Nyssaceae (Sour gum family) FAC
Quercus alba L. White oak QUAL Fagaceae (Beech family) FACU-
Quercus bicolor Willd. Swamp white oak QUBI Fagaceae (Beech family) FACW+
Quercus rubra L. Red oak QURU Fagaceae (Beech family) FACU-
142
Appendix B
SAS Programs
Example of mean calculations and t-tests used throughout Chapters 2, 3, and 4
DM 'LOG;CLEAR;OUTPUT;CLEAR';
OPTIONS LS=80 PS=75 PAGENO=1;
data soil;
input Plot$ Site$ no2no3ppm nh3ppm;
datalines;
<Insert data here>;
proc sort data=soil;
by param site;
quit;
proc ttest data=soil;
class site;
var mmean;
by param;
quit;
proc means data=soil n mean stderr;
by param site;
var mmean;
output out=soilmeans mean=soilmean stderr=soilse;
quit;
143
proc print data=soilmeans;
quit;
144
Example of Tukey-Kramer Procedure used in Chapter 3
DM 'LOG;CLEAR;OUTPUT;CLEAR';
OPTIONS LS=90 PS=100 PAGENO=1;
data litter;
input site$ plot$ date$ Species$ Litterg day$ Npct;
classday=day;
datalines;
<insert data here>;
TITLE1 'Npct';
title2 'Full model';
proc mixed data=litter;
class site species;
model Npct = site species site*species/ddfm=kr outp=resids;
lsmeans site species site*species/ adjust=tukey diff=all cl;
ods output lsmeans=lsmean1;
ods listing exclude diffs; ods output diffs=diff1;
ods output tests3=stat1;
quit;
proc sort data=litter;
by site;
quit;
title2 'by site';
proc mixed data=litter;
145
class site species;
model Npct = site species site*species/ddfm=kr outp=resids;
by site;
lsmeans species/pdiff adjust=tukey diff=all cl;
ods output lsmeans=lsmean1;
ods listing exclude diffs; ods output diffs=diff1;
ods listing exclude lsmeans;
ods output tests3=stat1;
%include 'a:pdmix800.sas';
%pdmix800(diff1,lsmean1,alpha=.05,sort=yes);
quit;
146
Example of repeated measures analysis used in Chapter 3 for determination of
concentration changes in tree leaf litter during the period of senescence
DM 'LOG;CLEAR;OUTPUT;CLEAR';
OPTIONS LS=80 PS=75 PAGENO=1;
data litter;
input site$ plot$ Date$ Nkgperha Litterg day;
classday=day;
datalines;
<insert data here>;
proc mixed data=litter;
class site day plot;
model Nkgperha = site day site*day /ddfm=satterth outp=resids;
random plot;
lsmeans site|day/adjust=tukey diff=all cl slice=day;
ods output lsmeans=lsmean1;
ods listing exclude diffs; ods output diffs=diff1;
ods output tests3=stat1;
quit;
proc plot data=resids vpercent=50;
plot resid*pred/vref=0;
quit;
data resids;
set resids;
147
aresid=ABS(resid);
run;
proc corr spearman data=resids;
var aresid pred;
quit;
proc univariate data=resids plot normal;
var resid;
quit;
proc print data=lsmean1;
quit;
proc print data=diff1;
quit;
proc print data=stat1;
quit;
148
Example of PROC MIXED procedure used to analyze the factorials in the
fertilization experiments in Chapters 4 and 5;
DM 'LOG; CLEAR; OUTPUT; CLEAR';
OPTIONS LS=90 PS=100 PAGENO=1;
data syfofert;
INPUT Ntrt$ Ptrt$ rep$ N C bio;
DATALINES;
<insert data here>;
proc mixed data=syfofert;
class Ntrt Ptrt rep;
model N = Ntrt Ptrt Ntrt*Ptrt/ddfm=satterth outp=resids;
lsmeans Ntrt Ptrt Ntrt*Ptrt/ adjust=tukey diff=all cl;
ods output lsmeans=lsmean1;
ods listing exclude diffs; ods output diffs=diff1;
ods output tests3=stat1;
quit;
proc plot data=resids vpercent=50;
plot resid*pred/vref=0;
quit;
data resids;
set resids;
aresid=ABS(resid);
run;
149
proc corr spearman data=resids;
var aresid pred;
quit;
proc univariate data=resids plot normal;
var resid;
quit;
proc print data=lsmean1;
quit;
proc print data=diff1;
quit;
proc print data=stat1;
quit;
150
Bibliography
Aerts, Rien. 1999. Interspecific competition in natural plant communities: Mechanisms,
trade-offs and plant-soil feedbacks. Journal of Experimental Botany 50, no.
330:29-37.
Anderson, Wendy B. and William G. Eickmeier. 2000. Nutrient resorption in Claytonia
virginica L.: Implications for deciduous forest nutrient cycling. Canadian
Journal of Botany 78:832-839.
Angier, Jonathan. 2001. Hydrologic controls on nitrate-N removal in a riparian system.
Hydrologic controls on nitrate-N removal in a riparian system, University of
Maryland, Ann Arbor, MI.
Angier, Jonathan T., Gregory W. McCarty, Clifford P. Rice, and Krystyna Bialek. 2002.
Influence of a riparian wetland on nitrate and herbicides exported from an
agricultural field. Journal of Agricultural and Food Chemistry 50:4424-4429.
Barbour et al. 1999. Terrestrial plant ecology. 3rd ed. Menlo Park, CA:
Benjamin/Cummings, Addison Wesley Longman, Inc.
Bedford, Barbara L., Mark R. Walbridge, and Allison Aldous. 1999. Patterns in nutrient
availability and plant diversity of temperate North American wetlands. Ecology
80, no. 7:2151-2169.
Bischoff, Joseph M., Paul Bukaveckas, Myron J. Mitchell, and T. Hurd. 2001. N storage
and cycling in vegetation of a forested wetland: Implications for watershed N
processing. Water, Air, and Soil Pollution 128:97-114.
Boerner, R. E. J. 1984. Foliar nutrient dynamics and nutrient use efficiency of four
deciduous tree species in relation to site fertility. Journal of applied ecology 21
151
:1029-1040.
Brenner, F. J. and J. J. Mondok. 1995. Nonpoint source pollution potential in an
agricultural watershed in northwestern Pennsylvania. Water Resources Bulletin
31, no. 6:1101-1112.
Cadenasso, M. L., M. M. Traynor, and S. T. A. Pickett. 1997. Functional location of
forest edges: gradients of multiple physical factors. Canadian Journal of Forest
Research 27:774-782.
Carpenter, S. R., N. F. Caraco, D. L. Correll, R.W. Howarth, A. N. Sharpley, and V. H.
Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen.
Ecological Applications 8, no. 3:559-568.
Casey, Ryan E., M. D. Taylor, and Stephen J. Klaine. 2001. Mechanisms of nutrient
attenuation in a subsurface flow riparian wetland. Journal of Environmental
Quality 30, no. 1732-1737.
Chambers, J. M., T. J. Wrigley, and A. J. McComb. 1993. The potential use of wetlands
to reduce phosphorus export from agricultural catchments. Fertilizer Research 36
:157-164.
Chapin, F. S. III and L. Moilanen. 1991. Nutritional controls over nitrogen and
phosphorus resorption from Alaskan birch leaves. Ecology 72:709-715.
Claridge, Kevin and Scott B. Franklin. 2002. Compensation and plasticity in an invasive
plant species. Biological Invasions 4:339-347.
Cooper, A. B. 1990. Nitrate Depletion in the Riparian Zone and Stream Channel of a
Small Headwater Catchment. Hydrobiologia 202:13-26.
Del Arco, Jose, M. Alfonso Escudero, and M. Vega Garrido. 1991. Effects of site
152
characteristics on nitrogen retranslocation from senescing leaves. Ecology 72, no.
2:701-708.
Eastman, John. 1995. The book of swamp and bog: Trees, shrubs, and wildflowers of
Eastern freshwater wetlands. 1st ed. (Mechanicsburg, PA: Stackpole Books).
Fenn, L. B., G. R. Gobran, and G. I. Agren. 1991. Nitrogen changes in forest litter after
fertilization with calcium-urea. Soil Science Society of America Journal 55, no.
509-514.
Gilliam, J. W. 1994. Riparian wetlands and water quality. Journal of Environmental
Quality 23:896-900.
Gleason, Henry A. and Arthur Cronquist. 1991. Manual of Vascular Plants of the
Northeastern United States and Adjacent Canada. 2nd ed. New York: New York
Botanical Garden Pr Department.
Gusewell, S., W. Koerselman and J.T.A. Verhoeven . 2003. N:P ratios as indicators of
nutrient limitation for plant populations in wetlands. Ecological Applications
13:372-384.
Hubbard, R. K. et al. 1999. Nutrient uptake and growth response of six wetland/riparian
plant species receiving swine lagoon effluent. Transactions of the American
Society of Agricultural Engineers 42, no. 5:1331-1341.
Joyce, Chris. 2001. The sensitivity of a species-rich flood-meadow plant community to
fertilizer nitrogen: the Luznice river floodplain, Czech Republic. Plant Ecology
155:47-60.
Kao, Jenny T., John E. Titus, and Wei-Xing Zhu . 2003. Differential nitrogen and
phosphorus retention by five wetland plant species. Wetlands 23, no. 4:979-987.
153
Killingbeck, Keith T. 1996. Nutrients in senesced leaves: Keys to the search for potential
resorption and resorption proficiency. Ecology 77:1716-1727.
Killingbeck, Keith T. and Ryan Tainsh. 2002. Does leaf size influence resorption of
nutrients from senescing leaves? Northeastern Naturalist 9, no. 2:213-220.
Ito, K., Y. Onda, T. Sato, Y. Abe, and M. Uemura . 2003. Structural requirements for the
perception of ambient temperature signals in homeothermic heat production of
skunk cabbage (Symplocarpus foetidus). Plant, Cell and Environment 26:783-
788.
Kirby, Robert M., Earle D. Matthew, and Moulton A. Bailey. Soil Survey of Prince
Georges County, Maryland. 1967. Washington, D.C., U.S. Government Printing
Office.
Koerselman, Willem and Arthur F. M. Meuleman . 1996. The vegeation N:P ratio: a new
tool to detect the nature of nutrient limitation. Journal of Applied Ecology
33:1441-1450.
Koutroubas, S. D., D. S. Veresoglou, and A. Zounos. 2000. Nutrient use efficiency as a
factor determining the structure of herbaceous plant communities in low-nutrient
environments. Journal of Agronomy and Crop Science 184:261-266.
Kozlowski, Theodore. 2002. Physiological-ecological impacts of flooding ono riparian
forest ecosystems. Wetlands 22, no. 3:550-561.
Kuo, S. 1996. Extraction with dilute concentrationn of strong acids. In Methods of Soil
Analysis Part 3- Chemical Methods, SSSA Book Series 5. (Madison, Wisconsin:
Soil Science Society of America, Inc.).
Lockaby, B. G. and W. H. Conner. 1999. N:P balance in wetland forests: Productivity
154
across a biogeochemical continuum. The Botanical Review 65, no. 2:171-185.
MDE. 2005.  ?Drought press releases,? Available from 
http://www.mde.state.md.us/Water/Drought/droughtReleases.asp. May 1, 2005.
Mitsch, William J. and James G. Gosselink. 1993. Wetlands. 2nd ed. New York: John
Wiley & Sons.
Morris, J. T. 1991. Effects of nitrogen loading on wetland ecosystems with particular
reference to atmospheric deposition. Annual Review of Ecology and
Systematics 22:257-279.
Mulvaney, R. L. 1996. Nitrogen - Inorganic Forms . In Methods of Soil Analysis Part 3-
Chemical Methods, Vol. 5, SSSA Book Series 5. (Madison, Wisconsin: Soil
Science Society of America, Inc.).
NOAA National Climatic Data Center. Climate Data for Beltsville, Maryland. Available
from http://lwf.ncdc.noaa.gov/oa/ncdc.html. July 14, 2004.
Palone, Roxane S. and Albert H. Todd eds. Chesapeake Bay Riparian Handbook: A
Guide for Establishing and Maintaining Riparian Forest Buffers. 1997. Radnor,
PA, USDA Forest Service.
Peterjohn, William T. and David L. Correll. 1984. Nutrient dynamics in an agricultural
watershed observations on the role of a riparian forest. Ecology 65, no. 5:1466-
1475.
Richardson, J. L. and M. J. Vepraskas. 2001. Wetland soils: genesis, hydrology,
landscapes, and classification. Boca Raton: Lewis Publishers.
Roberts, Mark R. and Frank S. Gilliam. 1995. Patterns and mechanisms of plant diversity
in forested ecosystems: Implications for forest management. Ecological
155
Applications 5, no. 4:969-977.
Schippers, Peter and Han Olff. 2000. Biomass partitioning, architecture and turnover of
six herbaceous species from habitats with different nutrient supply. Plant Ecology
149:219-231.
Schr?der, J. J., D. Scholefield, F. Cabral, and G. Hoffman. 2004. The effects of nutrient
losses from agriculture on ground and surface water quality: the position of
science in developing indicators for regulation. Environmental Science & Policy
7:15-23.
Seymour, Roger S. and Amy J. Blaylock. 1999. Switching off the heater: Influence of
ambient temperature on thermoregulation by eastern skunk cabbage
Symplocarpus foetidus. Journal of Experimental Botany 40, no. 338:1525-1532.
Silvan, Niko, Harri Vasander, and Jukka Laine. 2004. Vegetation is the main factor in
nutrient retention in a constructed wetland buffer. Plant and Soil 258:179-187.
Stone, K. C., P. G. Hunt, M. H. Johnson, and T. A. Matheny. 1998. Nitrate-N distribution
and trends in shallow groundwater on an Eastern coastal plains watershed.
Transactions of the American Society of Agricultural Engineers 41, no. 1:59-64.
Tessier, Jack T. and Dudley J. Raynal. 2003. Use of nitrogen to phosphorus ratios in plant
tissue as an indicator of nutrient limitation and nitrogen saturation. Journal of
Applied Ecology 40:523-534.
Tilman, D. 1985. The resource-ratio hypothesis of plant succession. American Naturalist
125, no. 6:827-852.
Todd, R., R. Lowrance, O. Hendrickson, L. Asmussen, R. Leonard, J. Fail, and B.
Herrick. 1984. Riparian vegetation as filters of nutrients exported from a coastal
156
plain agricultural watershed. Bioscience 34, no. 6:374-377.
USDA/NRCS National Plant Data Center. 2004. "The PLANTS Database v. 3.5.",
Available from http://plants.usda.gov.
Verhoeven, J. T. A., D. F. Whigham, M. van Kerkhoven, J. O'Neill, and E. Maltby. 1994.
Comparative study of nutrient-related processes in geographically separated
wetlands: towards a science base for functional assessment procedures. In Global
wetlands: Old World and New, (Amsterdam, The Netherlands: Elsevier Science
B.V.).
Vermeer, J. G. and F. Berendse. 1983. The relationship between nutrient availability,
shoot biomass and species richness in grassland and wetland communities.
Vegetatio 53, no. 2:121-126.
Vitousek, P. M. 1982. Nutrient cycling and nutrient use efficiency. American Naturalist
119:553-572.
Woodwell, G. M. 1974. Variation in the nutrient content of leaves of Quercus alba,
Quercus coccinea, and Pinus rigida in the Brookhaven Forest from bud-break to
abscission. American Journal of Botany 61, no. 7:749-753.
Xiong, Shaojun, Christer Nilsson, and Mats E. Johansson. 2001. Effects of litter
accumulation on riparian vegetation: Importance of particle size. Journal of
Vegetation Science 12, no. 231-236.