Page 1
Influence of Biogenic Silica from Terrestrial Vegetation on
Riverine Systems and Diatom Evolution
by
Beata Opalinska
A thesis submitted in conformity with the requirements
for the degree of Masters of Applied Science
Department of Earth Science University of Toronto
© Copyright by Beata Opalinska 2014
Page 2
All rights reserved
INFORMATION TO ALL USERSThe quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscriptand there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Microform Edition © ProQuest LLC.All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.789 East Eisenhower Parkway
P.O. Box 1346Ann Arbor, MI 48106 - 1346
UMI 1570576Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
UMI Number: 1570576
Page 3
ii
Influence of Biogenic Silica from Terrestrial Vegetation on
Riverine Systems and Diatom Evolution
Beata Opalinska
Masters of Applied Science
Department of Earth Sciences
University of Toronto
2014
Abstract
Presently within the scientific literature no terrestrial biogenic silica models exist that compare
by magnitude, processes transporting silica. Change in vegetation type has the potential to alter
dissolved concentrations of Si in rivers and ultimately the oceans. Diatoms greatly depend on Si
concentrations for growth, and as a result land cover change may have influenced onset diatom
radiation during the Cenozoic. To expand our understanding of this cycle, a terrestrial biogenic
silica model is proposed. This model accounts for biogenic silica production, dissolution and
leaching through soils, as well as providing estimates for annual silica soil storage. A case study
performed using the constructed biogenic silica model, showed an increase in oceanic DSi
concentration during the Miocene (period of diatom diversification). However, this increase does
not appear to have been sufficient to trigger global diatom radiation, suggesting multiple
geographically isolated locations for this diversification.
Page 4
iii
Acknowledgements
Thank you to Professor S. A. Cowling for your assistance and guidance, as well as your ever
constant insight into the Walking Dead series plot. Thank you to my committee, S. Finkelstein,
U.G. Wortmann and C. Mitchell for your constructive input for this thesis. Thanks to Kimsa
Dinh, Katie Schmidt, Anna Phillips, Veronica DiCecco and Sara Rhodes for all your
motivational support and ice-cream/pie breaks. Vasa Lukich for your ability to keep me
entertained when writing was not exciting enough, particularly with your familiarity of tumblr
and Supernatural. Thanks to Gary Vinegrad for inspiring last minute panic and Jessica Arteaga
for skating it out ;). Thanks to the physics gang (Josh Guerrero, Bob Tian, Bruno Opsenica, Sean
Langemeyer and Eric Goldsmith) for overwhelming me with learning new boredgames (ha!) and
fluid mechanics. Thanks to all of the 2013 and 2014 graduate students for your constant
amusement and friendship. Finally, thanks to my family who let me crash with them all of my
life, for those amazing lunches packed by my mom and clothes stolen from my sister.
Page 5
iv
Table of Contents
List of Tables……………………………………………………………………………………... vii
List of Figures…………………………………………………………………………………….. ix
List of Abbreviations……………………………………………………………………………... xi
List of Appendices……………………………………………………………………………...... xii
Chapter 1 Introduction……………………………………………………………………………. 1
1.1 Terrestrial Sphere……………………………………………………………………... 2
1.1.1 Sources of Terrestrial Biogenic Silica…………………………………..... 3
1.1.2 Benefits of using Biogenic Silica…………………………………………. 4
1.1.3 Soil Silica Storage………………………………………………………… 5
1.1.4 Marine and Terrestrial Silica Dissolution Kinetics……………………….. 7
1.1.5 Ecosystem Mass Balance…………………………………………………. 8
1.2 Aquatic Sphere……………………………………………………………………...... 9
1.2.1 Marine Silica Cycle……………………………………………………….. 9
1.2.2 Diatoms, Frustule Formation and Dissolution……………………………. 10
Chapter 2 Model and Materials…………………………………………………………………… 13
2.1 Terrestrial Biogenic Silica Model……………..……………………………………… 13
2.1.1 Production Reservoir………….…………………………………………. 14
2.1.2 Dissolution Flux………………………….………………………………. 14
2.1.3 Leaching Coefficient...………………………………………..…………. 15
2.2 Model Materials…………………………………………………………………........ 16
2.2.1 Gauge Selection…………………………………………………………… 16
2.2.2 Drainage Area Extraction…………………………………………………. 16
2.2.3 Land Cover………………………………………………………………… 17
2.2.4 Soils……………………………………………………………………….. 17
2.2.5 Precipitation and NPP Data……………………………………………….. 18
Chapter 3 Result and Model Validation…………………………………………………………... 22
3.1 Watershed Characteristics…………………………………………………..………… 22
Page 6
v
3.1.1 Watershed Productivity…………………………………………………… 22
3.1.2 Precipitation and Discharge…………………………………………......... 23
3.2 Watershed Silica Fluxes…………………………………………………………........ 24
3.2.1 Biogenic Silica Fixation Flux……………………………………...……… 24
3.2.2 Biogenic Silica Dissolution……………………………………………….. 24
3.2.3 Biogenic Silica Storage Reservoir………………...……………...……… 25
3.3 Biogenic Silica Riverine Flux……………………………………….……………….. 26
3.3.1 Leaching………………………………………………………………….. 26
3.3.2 Riverine DBSi Estimation……………………………………………...... 26
3.3.3 Riverine DBSi Seasonal Variation……………………………………….. 27
3.3.4 Soil Influence…………………………………………………………….. 27
3.4 General Trends in Riverine Biogenic Fluxes…………..……………………………. 28
Chapter 4 Discussion……………………………………………………………………………. 37
4.1 Biogenic Si Contributions………………………………………………………....... 37
4.2 Conifer Anomaly……………………………………………………………………. 37
4.3 Phytolith Dissolution……………………………………………………………....... 38
4.3.1 Surface Area Size and Dissolution……………………………………….. 39
4.3.2 Aluminum Induced Reduction in Dissolution…………………………..... 39
4.3.3 Influence of Soil Acidity (pH)….………….……………………..…….... 40
4.4 Soil Silica Transportation to Streams……………………………………………...... 41
4.5 Wetland Silica Retention…………………………………………………………….. 42
4.6 Regional Implications……………………………………………………………….. 43
4.7 Vegetation Influence on River BSi………………………………………………….. 44
4.8 Sources of Errors……………………………………………………………………. 45
Chapter 5 Case Study………………………………………………………………………….... 48
5.1 Global Oceanic Biogenic Silica Input………………………………………………. 48
5.2 Methods……………………………………………………………………………… 49
Page 7
vi
5.2.1 Oceanic Si Estimation……………………………………………………. 49
5.2.2 Paleo-Land Cover Distribution…………………………………………… 50
5.3 Results and Discussion………………………………………………………………. 50
5.3.1 Eocene to Pliocene Land Cover Change…………………………………. 50
5.3.2 Eocene to Pliocene Oceanic Si Change…………………………………... 51
5.3.3 What this mean for Diatom Radiation……………………………………. 52
5.3.4 What this means for Grasslands as an Instigator…………..……………… 53
5.3.5 Silica Retention…………..………………………………………………. 53
5.4 Global Impact………………………………………………………………………... 54
Chapter 6 Conclusions and Future Direction of TBSi Cycles………..…………………………. 59
References……………………………………………………………………………………….. 61
Appendix I……………………………………………………………………………………….. 76
Appendix II...……………………………………………………………………………………. 86
Page 8
vii
List of Tables
Table 1: Soil silica concentrations for various soil types, pH values and land covers. SiO2 is
dissolved silica.……………………..…………………………………………………....12
Table 2: Modeled Si fluxes, pools and rate constants for terrestrial systems with respective
calculations and variables. LF is the leaching factor, DF is the dissolution factor, NPP is
net primary productivity, %BSi is percent biogenic silica as net dry weight, SiD is the
average annual DSi, F is the soil water flow, mp is phytolith mass, SSA is the phytolith
specific surface area.………………………………………………………………..……19
Table 3: Phytolith specific surface area (SAA) and mass used for grasslands, wetlands,
coniferous and deciduous forests. …………………….…………………………………19
Table 4: Net primary productivity (NPP) and percentage biogenic silica net dry weight (%BSi)
of total plant weight for four analyzed land cover types………………………………..29
Table 5: Average catchment parametric values used for the calculation of dissolved silica (DSi)
fluxes for the four land cover types analyzed.........……………………………………...29
Table 6: Annual biogenic silica fixation rates from literature for wetland, grassland, coniferous
and deciduous ecosystems.…………………………………………………………..…..30
Table 7: Annual biogenic silica dissolution rates from literature for grassland, coniferous and
deciduous ecosystems.……………………………………………………………......….31
Table 8: Annual biogenic silica soil storage for wetland, grassland, coniferous and deciduous
ecosystems.……………………………………………………………………..………..32
Table 9: Annual dissolved silica flux from wetland, grassland, coniferous and deciduous
ecosystems.…………………………….………………………………………………...33
Table 10: Annual dissolved biogenic silica (DBSi) fluxes estimated using seasonality and
calculated leaching rates, in relation to annual dissolved silica (DSi) flux for four
ecosystems.……………………………………………………………………...……….34
Page 9
viii
Table 11: Modeled average catchment fluxes for each land cover type analyzed and corresponding
coefficients………………………………………………………………………………………..35
Table 12: Global area coverage (in ha) by various land cover types from the Eocene to
Pliocene…………………………………………………………………………………..56
Table 13: Total global flux of dissolved biogenic silica ( ) to oceans from the Eocene to
Pliocene and resulting oceanic biogenic silica concentrations (Tsi). Using ocean a
volume of 1.3 billion km3………………………………………………………………..56
Page 10
ix
List of Figures
Figure 1: Dissolution rates of phytoliths and quartz as a function of pH. Taken from Fraysse et
al., 2006.……………………..…………………………………………………………...12
Figure 2: Schematic of terrestrial biogenic silica model. Boxes in blue reflect dissolved silica,
boxes in white reflect silica in solid state. Dotted boxes refer to dominating processes
which influence the fluxes in the direction of the arrows. SAA refers to specific surface
area………………………………………………………………………………..……...20
Figure 3: Map of the United Sates of American showing point locations of the twenty-six gauges
studied and corresponding land cover. …………………………………………………..21
Figure 4: Depiction of relationship between precipitation, discharge and drainage for the
twenty-six gauges analyzed. Showing decrease in discharge with decrease in drainage
area and precipitation…………………………..……………………...…………………35
Figure 5: Dissolved silica fluxes and annual precipitation relationship between four studied
land cover types depicting leaching coefficients (r2 values). A. Grasslands B.
Wetlands C. Coniferous forests D. Deciduous forests…………………………………...36
Figure 6: The near 1:1 ratio of predicted dissolved biogenic silica flux using leaching coefficients
and seasonal segregation………………………………………………………………....36
Figure 7: A. Depiction of the relationship between Ge/Si ratios during the growing season,
and winter. B. Relationships of 30
Si/28
Si isotopes during the growing and winder
seasons. After White et al., 2012……………..………………………………………….47
Figure 8: Relationship between annual silica production and export. Coniferous regions show
low fixation but high export, while grasslands show the reverse………………………..47
Figure 9: Reconstruction of Eocene (55 Mya) land cover type and distribution. From After and
Ree, 2006……………………………….………………………………………………..57
Page 11
x
Figure 10: Reconstruction of Oligocene (27 Mya) land cover type and distribution. After Fine
and Ree, 2006 and Lunt et al., 2007…………………………………………………..…57
Figure 11. Reconstruction of Miocene (11 Mya) land cover type and distribution. After Pound et
al., 2011………………………………………………………………………….……….58
Figure 12. Reconstruction of Pliocene (3 Mya) land cover type and distribution. After
Haywood et al., 2004.…………………………………………………..……….……….58
Page 12
xi
List of Abbreviations
Al: Aluminum
ATP: Adenosine Triphosphate
ASi: Amorphous Silica
TBSC: Biogenic Terrestrial Silica Cycle
DBSi: Dissolved Biogenic Silica
DSi: Dissolved Silica
GCM: Global Climate Model
Ge: Germanium
TBSi: Terrestrial Biogenic Silica
TSC: Terrestrial Silica Cycle
Na: Sodium
NPP: Net Primary Productivity
Si: Silica
Page 13
xii
List of Appendices
Appendix I Biogenic silica content of vegetation (%DW) belonging to grasslands, wetlands,
coniferous forests and deciduous forests………………………………………………...73
Appendix II Calculated terrestrial BSi model parameters for each catchment…………………..83
Page 14
1
Chapter 1
1.0 Introduction
The importance of silica in the terrestrial environment has been recognized since the late 1980’s,
but has just recently come of interest for biogeochemists. Prior to the 1980’s and even now, the
role of biogenic silica has been largely excluded from global continental cycles of carbon and
silicon. As a result, our current understanding of the biogenic terrestrial silica cycle (TBSC) is
limited. While several studies have attempted to describe and compare terrestrial biogenic silica
by magnitude, processes transporting silica and fluxes are neither well known nor quantified. In
contrast, the role of lithogenically derived silica has been well established within terrestrial and
marine environments. In nearly all marine silica models lithogenic silica is the only noted source,
and the quantity of published work pertaining to this silica attests to its dominance in this
subject. In order to expand our understanding of TBSCs, and minimize the discrepancy in what
is known between the two sources, a terrestrial silica model is proposed that can be applied for
several vegetation types.
Biogenic silica (BSi) reservoirs in terrestrial environments include living vegetation, soils and
rivers. Generally, these pools are mainly influenced by processes of deposition, dissolution and
leaching. Silica initially enters the biogenic terrestrial cycle as lithogenic silica and is converted
into biogenic forms by vegetation. Upon deposition, silica from vegetation is added to the soil
reservoir and undergoes dissolution (Collin et al., 2012). Changes in vegetation greatly influence
soil silica quantities through litterfall. As vegetation belonging to one land cover type shows
comparable biogenic silica values, land cover classes can be associated with specific production
rates. The dissolution of BSi in soils occurs at a much higher rate than inorganic silica, resulting
in a dissolved biogenic silica (DBSi) reservoir. This dissolved silica leaves the system through
leaching primarily by way of precipitation. Once leached from soils, the DBSi component in
addition to lithogenically derived dissolved silica enters rivers and subsequently oceans. Current
estimates of oceanic biogenic silica contributions are 1.1 Tmol Si year-1
(Treguer and De La
Rocha, 2012). This value is made up of two components, one comprised of freshwater diatom
silica and the other of silica reworked by vegetation. Many studies have found the flux of
Page 15
2
biogenic silica to oceans to be less than that of lithogenic yet significant, and integration of this
component into marine models would be beneficial. Understanding how this value changes with
changing terrestrial ecology could have severe implications for the amount of silica entering,
circulated and deposited within the marine system.
The applications of a terrestrial Si model are vast and its inclusion to current silica models would
greatly enhance our understanding of biogeochemical cycles; not to mention emphasize the
complex influence terrestrial ecology has on earth systems. One application of this model would
be to estimate global silica concentration changes from one geologic period to another using
change in land cover through deep time and global climate models (GCMs). This information
can be used to predict changes in oceanic biogeochemistry through time. In addition, this model
can inform us on oceanic silica conditions during diatom diversification/radiation events.
Diatoms are the focal organisms that use silica for corporeal functions and in photosynthesizing
provide a large portion of our breathable oxygen.
Due to the potential influence of terrestrial biogenic silica for biogeochemical cycles and
productivity in the oceans, it is necessary to quantify and model this poorly understood system.
In order to do so, research was conducted with objectives as follows:
(1) To create a simple terrestrial biogenic silica model
(2) To determine if land cover influences concentration of silica in riverine settings, and if so
(3) How would land cover changes influence global dissolved silica, and could it be used to
(4) Determine if changes in terrestrial ecology during the Oligocene triggered diatom
evolution/radiation
1.1 Terrestrial Sphere
A fair amount of information is available concerning silica in the terrestrial environment. For
instance, silica concentrations of various plants have been of interest since the late 1970’s, and
can now be compiled into a sizable database (Klein and Geis, 1978; Hodson et al., 2005). In
Page 16
3
addition, soil silica distribution concentrations are relatively abundant as are dissolved riverine
concentrations (Sommer et al., 2006). Recently, several attempts have been made to construct
mass balance reconstructions of silica for numerous environments, but often not all silica pools
are analyzed. This discrepancy makes our understanding of the terrestrial silica cycle (TSC) still
incomplete. To create a comprehensive model of silica movement, compilation of data from
these sources and quantification of processes is necessary. Silica housed in vegetation is the
primary source of biogenic silica within the terrestrial sphere, eventually deposited into soils
upon plant death. Once in soils, this silica undergoes dissolution, leaching into rivers and
movement into oceans.
1.1.1 Sources of Terrestrial Biogenic Silica
Vegetation, both terrestrial and aquatic, can be divided into two categories based on silica
accumulation. Accumulator species which contain >10 mg g-1
of BSi are considered enriched in
Si. Within angiosperms, species belonging to orders Poales, Saxifragales and Arecales
accumulate some of the highest values of BSi. Bamboo species generally have 13 to 23% BSi,
grasses 2 to 4.7% BSi and rice approximately 2% BSi, however these tend to range broadly
between species (Bezeau et al., 1966l; Collin et al., 2012). Different parts of plants can express
large variations in silica accumulation. For bamboo, approximately 58% of silica can be found in
leaves, 14% in branches, 17% in stems and 10% in roots (Ding et al., 2009). Non-accumulator
species contain <5 mg g-1
of BSi and include the majority of dicots (flowering plants), ferns and
conifers. Representatives from these taxa have very low amounts of biogenic silica in their
vegetative structures, 0.48% BSi in oaks and 0.13% BSi in pines (Geis, 1978; Klein and Geis,
1978). The reasons for variation in accumulation between species can be mainly attributed to the
ability of Si uptake by roots. These variable silica concentrations of plant species can be
averaged providing general biome Si accumulation rates (Carey and Fulweiler, 2012).
The differential use of silica in plants has been found to alleviate many stresses ranging from
predation to maintaining stem rigidity. The expenditure of silica for these functions has been
evolutionarily selected for, leading away from the use of carbon materials (Raven, 1983).
Incorporating silica has been found to be energetically cheaper than other leading structural
Page 17
4
materials, particularly lignin. Through the use of Si stoichiometry and Si presence in cell walls of
plants, it has been found that only one adenosine triphosphate (ATP) is required per Si
conversion. No further energy is required for metabolism and transport of Si from point of
entrance to precipitation. Thus one molecule of SiO2 can be precipitated in cell walls at the cost
of one ATP (Raven, 1983). In contrast, on a weight basis, the energetic cost of incorporating
lignin is 27 times that of incorporating 1 g of SiO2. When converting into volumes, more
pertinent for rigidity, 1 g of lignin is 20 times more costly and polysaccharides 10 times.
Although silica is more efficient to metabolize, it is not as common a structural material in the
plant kingdom and this is believed to be a result of available SiO2 depletion in soils (Cooke and
Leishman, 2011)
There are two leading mechanisms thought to be responsible for soluble silica uptake by plants.
The two processes include active transport of silicic acid by metabolic processes and passive,
nonselective flow of silicic acid from soil water through transpiration. Following up take, silica
is moved from cortical cells to the xylem, mediated by energy dependent transport processes. In
the xylem, silicic acid polymerizes to form silica gel. This polymerization occurs when silicic
acid concentrations exceed a threshold value of 2 mM (Ma, 2006). In the shoots of plants silicic
acid is further concentrated through transpiration. Silica is deposited as a layer in the space
directly below the cuticle layer in leaves, forming a cuticle Si double layer as seen in Si
accumulator species like rice (Ma, 2006). In the leaf blades two types of silica forms can be
found: silica cells and silica bodies. Silica cells include free floating unbound biogenic silica,
while silica bodies consist of phytoliths and opal which are precipitated forms of silica cells.
1.1.2 Benefits of using Silica
Deposition of silica in both phytolith and unbound forms acts to benefit plants from both biotic
and abiotic stresses. High concentrations of silica in rice, strawberry plants, barley and
muskmelon have been found to supress the effect of fungal disease (Datnoff, et al., 2007; Kanto
et al., 2006; Zeyen, 2002; Fauteux et al., 2011). Two hypotheses are available for explaining this
silica-based resistance to disease. One explanation is that the Si deposited beneath the cuticle
player acts as a physical barrier preventing infiltration of fungal pathogens and making the
Page 18
5
tissues less susceptible to enzymatic degradation. An alternate explanation is that Si enhances the
production of phytoalexin, an antimicrobial chemical (Ma and Miyake, 2001). Furthermore,
silica accumulation acts to increase abrasiveness of foliage deterring herbivory through tooth
enamel reduction and increases energy required for digestion (Gali-Muhtasib et al., 1992;
Massey et al., 2006; Massey et al., 2007; Garbuzov et al., 2011).
Silica has also been seen to alleviate physical stresses caused by radiation, water stress, and high
winds. The incorporation of silica into stems promotes wall thickening by increasing the size of
vascular bundles, thereby increasing rigidity of stalks preventing irreparable damage (Ma and
Miyake, 2001; Casler and Jung, 2006; Hill and Pickering, 2009). The Si-cuticle double layer that
is formed upon deposition of Si bodies is seen to significantly reduce transpiration allowing Si
accumulating plants to better cope with water-stressed conditions (Ma et al., 2001). The decrease
in transpiration also aids plants grown under saline conditions by blocking the pathway through
which sodium (Na) is absorbed (Yeo et al., 1999). In addition to these abiotic stresses, Si
accumulation also assists with preventing heavy metal toxicity particularly involving manganese,
iron and zinc, but also aluminum. In the case of these metals Si accumulation leads to reduced
uptake, encourages homogenous distribution, and modifies cation binding properties (Okuda and
Takahashi, 1962; Horst and Marschner, 1978; Horst et al., 1999).
1.1.3 Silica Soil Storage
Typically, soils are the main and largest medium in which terrestrial processes facilitate both
chemical and biological interactions. Soil silica goes through a process of formation, deposition,
dissolution and leaching. Because silica is not synthesized by biological processes, vegetation
must accumulate Si from a source and subsequently convert it to useable forms. The soil silica
pool includes two forms of silica, one being mineral and the other amorphous (ASi). From these
two groups, relative contributions are not well quantified making an analysis of this system
challenging. The mineral pool is comprised of two silicate forms, primary minerals which are
inherited from parent materials and secondary minerals that are developed through soil
formation. Crystalline silicates include quartz, plagioclase, clay minerals and feldspar while the
amorphous forms are mainly dominated with phytoliths and biogenic silica converted by plants.
Page 19
6
Silica concentrations in soils can be seen to vary widely ranging from < 1 to 45% dry weight
(Sommer et al., 2006). Silica inputs into the soil system primary include dust or aeolian materials
and litter fall into topsoils. Soil phytolith distributions vary with depth, often reflecting a
negative asymptotic curve. In several grassland systems, the top 20 cm of soil reflects over 60%
of the total soil phytolith assemblage (Blecker et al., 2006). Translocation also occurs from
surface sources displacing phytoliths further down the soil profile. After rainfall events,
mineralization, increase in acidity and formation of organic compounds in top soils occurs.
Organic compounds react with soil minerals resulting in high concentrations of Si as well as Al
and Fe. This process occurring in the topsoils, where phytolith restitution occurs, results in high
dissolved silica concentrations closer to the surface through the soil profile. However, dissolution
of lithogenic silica leads to an increase of dissolved silica at depth, matching dissolved silica
quantities at the surface (Gerard et al., 2002).
Following deposition and formation, silica in soils undergoes a process of dissolution. Solubility
of quartz and amorphous silica differs greatly, 1.8 to 2 mM Si and 0.10 to 0.25 mM Si
respectively. This is attributed to a higher density of tetrahedral structure in quartz silica and
crystal order (Drees et al., 1989). Within plant species dissolution rates of phytoliths (a part of
ASi) differ based on sorption of Al and other metals (Fe3+
and Zn2+
). For instance, the solubility
of pine phytoliths is several times lower than beech on account of higher Al substitution seen in
pine (Hodson and Evans, 1995). Silica in soils can appear as either silicic acid and/or an ionized
solution [Si(OH)3O-]. Soil silica concentrations can vary from 0.03 to 0.6 mM (Epstein, 1994). In
the case of acidic podzol soils, clay breakdown can lead to the mobilization of Si increasing
concentrations (Sommer et al., 2006; Frank 1993). Dissolution rates also differ with the presence
or absence of vegetation where rates are lower without plants (Hinsinger et al. 2001).
Vegetation also influences silica concentrations in soils through weathering and absorption.
Terrestrial plants affect silicate mineral weathering through changing soil temperatures,
preventing erosion, altering pH through organic acid production, modifying soil solution
concentrations and water dynamics (Drever, 1993). Although vegetation exerts process both
promoting and hindering weathering, the net influence is to increase weathering. Studies have
found that weathering and nutrient release rates increase by a factor of 2 to 5 with the presence of
Page 20
7
vegetation (Moulton and Berner 1998, Hinsinger et al. 2001). Silica released during weathering
is recycled and forms a component within soils where DSi is available for plant-uptake. If a
region is characterized by BSi accumulator species then silica in soils is significantly reduced
until deposition of foliage (Meunier et al., 1999). Silica concentrations found in soils are greatly
influenced by overlying plant material, pH and soil type (Table 1). Various soil types are able to
display different Si concentrations even with similar land cover and pH due to the influence of
underlying parent material (Sommer, 2006).
1.1.4 Marine and Terrestrial Silica Dissolution Kinetics
Understanding the dissolution kinetics of biogenic silica is essential as this process will dictate
silica leaving soil systems. Several equations have been derived predicting silica dissolution rates
following the general form given by Lasaga et al. (1984);
∏ (1)
where, is the dissolution rate (mol cm-2
s-1
), k is the rate coefficient of the dissolution reaction,
A is the surface area (cm2
g-1
), Ea is the activation energy, R is the universal gas constant, T is
temperature (K), a is the pH dependent term, and Gr is the Gibbs free energy of reaction.
Following the dissolution reaction, the kinetic energy possessed due to motion varies from 0.09
to 60 mol g-1
h-1
for cool waters and from 0.65 to 450 mol g-1
h-1
for warm waters (Rickert et
al., 2002). General dissolution rates for BSi in both freshwater and marine waters varies from 0.1
to 10.1 mol g-1
h-1
under constant abiotic conditions (Loucaides et al., 2008). Case specific
dissolution models can be viewed in Dove et al. (2007), Loucaides et al. (2008), and Fraysse et
al. (2008, 2009).
Silica dissolution models reveal rates to be greatly influenced by temperature, salinity and pH.
Dissolution of silica appears to occur at a faster rate in waters of higher temperature, and
similarly in environments of higher pH and salinity (Fraysse et al., 2006). A temperature rise
reflects increased energy available to initiate bond breakage from biogenic silica to silicic acid
and water. While an increase in pH leads to increased deprotonation of surface silanol bonds also
Page 21
8
resulting in bond breakage. The relationship between dissolution and pH can be expressed as a
negative parabolic function with a vertex centered at a pH of 3 to 5 depending on the silica
source (Figure 1). When analyzing phytolith BSi dissolution, the vertex occurs at a pH of 3,
while diatom derived biogenic silica dissolution is at a minimum at a pH of 5 (Greenwood et al.,
2001). As a result, dissolution rates of silica define three regions. For strong acidic solutions (pH
< 3) rates increase with , at 3 ≤ pH ≤ 5 rates are independent of pH, and at pH from 5 to 12
dissolution rates increase. Various studies have shown that dissolution rates of quartz and
amorphous silica increase 50 to 100 times with an increase in alkalinity (Van Cappellen and Qui,
1997; Dove et al., 2007).
1.1.5 Ecosystem Mass Balance
When reviewing literature regarding biogeochemical processes of TSCs, it is evident that there is
a paucity of mass balance reconstructions and no analytical models have been established. Of
ecosystems to be studied grasslands demonstrate the highest Si fixation rate ranging from 166 to
350 kg ha-1
yr-1
(Bartoli, 1983). Comparably, bamboo forests produce large quantities of BSi 97
to 138 kg ha-1
yr-1
(Meunier et al., 1999). Bamboo forests show inflated silica fixation rates a
result of rapid plant growth (averaging 3-10 cm day-1
) (David, 1984). Temperate deciduous and
coniferous forests display some of the lowest fixations rates, 27 kg Si ha-1
yr-1
and 8 kg Si
ha-1
yr-1
, respectively (Carnelli et al., 2001). On an annual basis the amount of Si taken up by
vegetation is equal to or less than Si deposition though litterfall. DSi that is returned to the soil
interface from vegetation can be taken up once again and forms a recoverable component of the
Si mass balance. This recycled component, equivalent to the biomass BSi, does not contribute to
leached DSi in rivers, and in fact delays DSi transport. In one forest site it was estimated that
80% of the DSi export was recycled through a deciduous ecosystem, compared to 20% for a
coniferous forest (Bartoli, 1983). Silica in soils is a function of biomass BSi, where higher
biomass BSi leads to increased soil Si. Ecosystem Si can be seen to range from 50, 000 kg ha-1
in
coniferous forests to 250, 000 kg ha-1
in grasslands (Bartoli, 1983; Blecker et al., 2006).
Terrestrial mass balance calculations of silica reveal that biogeochemical cycling occurring in
forested/grassland ecosystems is considerable. Export from these systems is relatively minute
Page 22
9
considering the volume of biogenic silica stored in soils, yet the main source of DSi delivered to
oceans. When looking at the balance of silica at a watershed scale, understanding Si pools and
pathways is necessary. In soils, silica goes through recycling by reabsorption via vegetation,
immobilization through plant retention, net deposition and finally leaching. Leaching is the
process by which silica moves through the soil profile, stimulated by precipitation. Several
studies have found that land cover indeed influences DSi concentrations in rivers, but the extent
of this relationship varies between ecosystems. Calculated relative influence factors for land
covers on observed Si fluxes varied between 0.041 for deciduous forests to 0.260 for wetlands
(Carey and Fulweiler, 2012). To further support this claim a study conducted by Song et al.
(2011) revealed that there is a significant difference in SiO2 concentrations for bamboo, mixed
forest and broadleaf watersheds. Concentrations reflected 120 × 10-6
mol L-1
, 40 × 10-6
mol L-1
and 65 × 10-6
mol L-1
for bamboo, mixed forest and broadleaf watersheds respectively.
1.2 Aquatic Sphere
In order to appreciate effects that the TBSC might have on the marine ecosystems, a general
review of silica in the oceans is given. Vegetation can influence marine DSi through mass
production or retention, or have no effect. Ultimately, 7.3 Tmol Si year-1
is exported into ocean
waters, which undergoes intense remineralisation by diatoms and deposition (Treguer and De La
Rocha, 2012).
1.2.1 Marine Silica Cycle
Our current understanding of the marine silica cycle is limited by our lack of knowledge
concerning biogenic silica inputs from rivers. However, our general understanding of other
source fluxes, circulation and deposition into the oceanic sphere is well supported by both
theoretical and physical evidence. Recent riverine estimates of current global biogenic silica
contributions are of 1.1 Tmol Si year-1
, while lithogenic contributions are of 6.2 Tmol Si year-1
(Treguer and De La Rocha, 2012). Silica also enters the marine cycle by means of groundwater,
sea floor weathering, aeolian and hydrothermal processes, adding approximately 3.6 Tmol Si
year-1
(Treguer et al., 1995). Once in the oceans the dissolved amorphous silica is used by
diatoms to synthesize skeletal structures and as a by-product produce biogenic silica. It is
Page 23
10
estimated that diatoms produce approximately 240 Tmol Si year-1
and resultantly account for
40% of marine primary productivity and 50% of organic carbon burial in marine sediments
(Nelson et al., 1995; Falkowski et al., 2004). Following biogenic silica production,
approximately 105 Tmol Si year-1
leaves surface waters. Of that, 6.3 Tmol Si year-1
is deposited
in costal and abyssal sediments, with the difference in fluxes recycled within the water column.
The overall residence time of silica in the oceans is estimated to be 10,000 years (Treguer and De
La Rocha, 2012), falling between that of nitrogen, < 3,000 years (Sacramento and Gruber, 2006),
and phosphorous, 30,000-50,000 years (Delaney, 1988). This value and the resident time relative
to biological uptake suggest that silica in the oceans is cycled approximately 24 times before
deposition to sea floor sediments (Treguer and De La Rocha, 2012).
1.2.2 Diatoms, Frustule Formation and Dissolution
Ultimately, the silica flux into oceans directly influences primary productivity. Ocean NPP is
highly dependent upon silica concentrations as diatoms which are large oceanic NPP contributors
metabolize silica intended for creating skeletal structures. Numerous studies have shown that the
concentration of silicic acid in aqueous environments acts as a regulating nutrient for diatom
dominance (Jorgensen, 1952). In particular, a study conducted by Egge and Aksnes (1992)
showed that a minimum requirement of 2 of dissolved silicic acid is necessary for diatom
dominance to attain 70% richness. For Cenozoic diatom evolution, this absolute requirement is
thought to have been catalyzed by some event that led to an increase of soluble silica in marine
ecosystems (Rabosky and Sorhannus, 2009). One hypothesized such event is the evolution and
expansion of grasslands that occurred concurrently with diatom radiation. Presently, the use of
silica by diatoms and other siliceous organisms such as sponges and radiolarians, has led oceans
to be ubiquitously undersaturated in silicic acid (Siever, 1991). Diatoms first appear in the fossil
record approximately 185 mya and in abundance 40 mya during the Eocene/Oligocene transition.
Before the evolution of siliceous plankton DSi was relatively abundant in seawaters with
concentrations near saturation. Presently, diatoms have depleted the oceans of Si where
concentrations are generally <10 at the surface and <160 in deep waters (Treguer and De
La Rocha, 2012).
Page 24
11
The uptake of silicic acid by diatoms can occur through one of many transporter genes
responsible for regulation to maintain supersaturation. Once within the organism, polymerization
of the silica occurs converting monosilicic acid to hydrated amorphous silica (general reaction
SiO2(s) + 2H2O = H4SiO4). This reaction is an overall thermodynamically favourable process.
Silica polymerization occurs within tracellular compartments, called silica deposition vesicles
(SDVs), which are bound by silicalemma converting aqueous silica into solid deposits. Not only
does the SDV play a role in polymerization, but once the silica has been formed, it also acts as a
mold by the cytoskeleton to form the final silicified profiles of frustules. Under silica limited
conditions most diatom species are unable to complete wall formation, inhibiting cell division
and growth. This explains why growth is more rapidly hindered under Si starvation as opposed to
other nutrients.
In addition to silicic acid limits on diatom metabolism, other nutrients and water-atmospheric
conditions, will determine the distribution of diatoms. Current diatom distributions have been
modeled and reflect diatom dominance in high and low latitudes and in equatorial and coastal
upwelling regions (Kamykowski et al., 2002; Gregg and Casey, 2007). Diatoms are typically
found in regions with plentiful nutrients (nitrogen, ammonium and iron), abundant light and in
cooler waters, this is believed to be a result of high maximum growth rates, related to the
efficiency of metabolizing silica (Gregg and Casey, 2007). Although diatoms can be seen to
dominate over other phytoplankton found in these zones, the persistence of diatoms can also be
greatly limited by alkalinity of the water. In waters of high pH, dissolution rates for BSi are
increased, however, whether this negatively influences diatoms through cell wall dissolution, or
positively influences them through regeneration of bioavailable DSi, is unknown (Lewin, 1961;
Ryves et al., 2006; Loucaides et al., 2008).
Page 25
12
Table 1. Soil silica concentrations for various soil types, pH values and land covers. SiO2 is
dissolved silica.
Soil Type pH Parent
Material Land Cover SiO2 (mg g
-1)
Podzol 3.7 – 3.9 Mica schist Coniferous 55
3
Podzol 3.12 – 4.6 Mica schist Deciduous broadleaf
Podzol 2.7 – 3.8 Sandstone Deciduous broadleaf 9 3
Luvisol 3.7 – 4.4 Loess Deciduous broadleaf 12 3
Regosol 7.0 Loess Deciduous broadleaf -
Vertisol 4.4 – 5.1 Claystone Deciduous broadleaf 18 3
Planosol 3.2 – 3.8 Gneiss Coniferous 6 3
Leptosol 7.2 Limestone Deciduous broadleaf -
Chernozem 5 – 8.4 Sedimentary Grassland 22 – 93 3
Histosols 6.5 – 7.5 1 Organic Peat Wetlands 2.3
2
1. Given and Miller, 1985 2. Struyf and Conley, 2009 3. Saccone et al., 2007
Figure 1. Dissolution rates of phytoliths and quartz as a function of pH. Taken from
Fraysse et al., 2006
Page 26
13
Chapter 2
2.0 Model and Materials
As of yet, no study has attempted to construct a terrestrial plant Si model predicting
concentrations of silica within reservoirs. Terrestrial mass balances are available for various land
covers, but belong to specific vegetation types and environmental conditions. This specificity
makes case generalizations and comparisons between regions challenging. To model how land
cover influences pools and fluxes of the Si cycle, several relationships are described using
theoretical approaches, and quantified using numerous data resources.
2.1 Terrestrial Biogenic Silica Model
The biogenic silica model in this study was constructed to reflect the movement of biogenic
silica through the terrestrial sphere, from phytolith to dissolved forms. The model constructed
emphasizes the transition of plant BSi to DSi within soils through the process of dissolution, and
subsequent leaching. Biogenic silica can be found in four reservoirs within the terrestrial system
(Figure 2). The first reservoir expresses biogenic silica that is found within vegetation, the
production of siliceous materials, which are deposited through litterfall and buried in soils. The
second reservoir of biogenic silica consists of phytoliths that are found within the soil annually
and do not exit the system through dissolution. The third reservoir consists of dissolved biogenic
silica that can be found in soils annually that does not leach from the system. Finally, the fourth
reservoir is made up of leached dissolved riverine biogenic silica that is eventually deposited into
the oceans. Seawater can be considered a fifth reservoir when including marine environments.
The dissolved biogenic flux of silica into rivers can be estimated using the equation;
= LF · DF (NPP · %BSi), (2)
where, LF is the leaching factor, DF is the dissolution factor, NPP is the net primary productivity
(kg ha-1
y-1
) and %BSi is the percent of biogenic silica found in plant tissues. The relationships
and calculations for all reservoir turnover and flux rates can be viewed in Table 2.
Page 27
14
2.1.1 Production Reservoir
Biogenic silica content data for vegetation was divided into four categories reflecting land cover
classes. Grassland, wetland, coniferous forest and broadleaf deciduous forest classes were
selected to represent broad regional ecosystems analogous to those of the Cenozoic.
Additionally, each of these classes is believed to influence dissolved riverine biogenic silica
concentrations differently. Biogenic silica content was estimated using;
BSipro = NPP · %BSi, (3)
where BSipro is production (kg ha-1
yr-1
), NPP is net primary productivity (kg ha-1
yr-1
) and %BSi
is biogenic silica content in plant tissues as dry weight. A biogenic silica concentration database
was constructed to determine potential silica inputs into the terrestrial cycle (Appendix I). This
database was limited to foliage silica and included only vegetation found within the United
States. The data was collected from and categorized by plants belonging to the four land cover
classes as mg kg-1
and then converted to percentage. To account for silica allocation in structures
not deposited annually (i.e. stems), forest %BSi was weighted to account for 30% of forest NPP
(Litton et al., 2007).
2.1.2 Dissolution Flux
The dissolution flux was calculated for each catchment (Appendix II) irrespective of initial
phytolith biogenic silica stored in soils assuming that the system is not limited by this silica
reservoir. The rate of phytolith dissolution was adapted from the equation developed by Fraysse
et al. (2009);
, (4)
where is the dissolved riverine silica concentration (mol l-1
), Q is the water percolation
through soils (L s-1
), msi is the mass of phytoliths (g), and S is the specific surface area of
phytoliths (cm2 g
-1). Water percolation was estimated using hydraulic conductivities for specific
soil types and integrated per area. The mass and specific surface area of larch, elm, and horsetail
phytoliths were taken from Fraysse et al. (2009), and used to represent conifer and broadleaf
Page 28
15
deciduous forests and grasslands/wetlands, respectively (Table 3). The dissolution factor as seen
in eqn. (2) is calculated as;
⁄ , (5)
where R is the dissolution flux (kg ha-1
yr-1
) and BSipro is the biogenic silica production flux (kg
ha-1
yr-1
). This constant is dependent upon vegetation class and its inclusion into the model
allows for correction of the amount of silica leaving terrestrial systems.
2.1.3 Leaching Coefficient and Factor
The leaching coefficient which reflects the movement of dissolved silica through soils and into
rivers was calculated using regression analysis of precipitation and DSi concentrations found in
river waters. Precipitation in this case acts as the moving mechanism and medium by which
silica is transported through soils. Dissolved silica concentrations were obtained from the USGS
Water-Quality data set (See 2.2.1 Gauge Selection) and converted into silica fluxes by
incorporating discharge and standardizing by drainage area. This linear relationship suggests a
constant of proportionality for DSi in waters that can be explained by leaching, and not through
diatom production or direct riverine substrate dissolution. The leaching factor in eqn. (1) is
calculated as;
⁄ , (6)
where DSi is dissolved silica (measured in rivers by the USGS) (kg ha-1
yr-1
), R is the dissolution
flux (kg ha-1
yr-1
), and LC is the leaching coefficient, described above. This constant is also
dependent upon vegetation class and further constrains predicted DBSi leaving systems.
Constants were analyzed and compared among and between soil types to discern geologic
influence. Due to limitation of data availability this relationship includes both amorphous and
mineral forms of silica as opposed to solely desired biogenic forms. To distinguish between the
two components dissolved silica data was separated into two periods, one from October to April
and the other from May to August. The October to April silica values represent mostly biogenic
silica inputs; this time period corresponds to the non-growing season when uptake from soils is
Page 29
16
minimized and organic soil horizons have added material. During these months weathering
processes also decline to a minimum reducing the contribution of mineral silicates. The May to
August period represents a time during which lithogenic silica dominates the DSi flux.
2.2 Model Materials
2.2.1 Gauge Selection
To study the influence of biogenic silica on riverine systems, dissolved silica data was collected
from twenty-six (26) gauges distributed across the U.S (Figure 3). Gauge data was obtained from
the U.S Geological Survey Water Quality Field/Lab sample database as dissolved silica in mg l-1
.
Each gauge represents a minimum of eight monthly observations per year to accurately estimate
annual average riverine dissolved silica content. For several gauges which expressed an
abundance of observations, monthly averages were calculated as well. Gauges were also
constrained by drainage area (1 to 500 sq. mi), proximity to urban developments such as cities,
and period of record (2005 to 2012). The twenty-six selected gauge locations were imported into
ArcGIS and used to predict drainage areas, subsequently used to extract land cover type, soil,
NPP and precipitation data for use in the terrestrial silica model.
2.2.2 Drainage Area Extraction
Once gauges exhibiting desired parameters were selected, they were viewed using the USGS
National Water Information System Map Viewer and exported as an ESRI shapefile and
imported into ArcGIS. To predict the drainage basin extent of each gauge, the hydrology based
spatial analyst toolset within ArcGIS was used. Digital elevation models (DEMs) for this
analysis were obtained from the USGS National Elevation Dataset and collected at 1 arc second
(30 m resolution) in a raster arcgrid format. Using the hydrologic analysis tools and acquired
DEMs, flow accumulation and direction was calculated to delineate watershed area that would
contribute to and influence dissolved silica concentrations at the gauge point locations.
Following drainage extent determination, areas were geographically overlaid with the physical
data files to determine corresponding dominate cover, soil types and average precipitation.
Page 30
17
2.2.3 Land Cover
For this study four land cover regions; grasslands, wetlands, coniferous forests, and broadleaf
deciduous forests, found within the United States of America were defined. To geographically
select gauges belonging to these cover types; the Land Cover database of North America for the
year 2000 was used. This dataset was generated by Natural Resource Canada and the U.S
Geological Survey for the Global Land Cover 2000 (GLC2000) project, implemented by the
Global Vegetation Monitoring Unit, Joint Research Centre of European Commission. For each
drainage the dominant cover type was determined based on at least 70% drainage area coverage.
The Land Cover database of North America was created using SPOT VEGETATION data for
the growing season in 2000 at a spatial resolution of 1 km. This data was subsequently converted
into a regional land cover product map, consisting of thirty-five (35) land cover classes based on
the modified Natural Vegetation Classification Standard (NVCS) used by the U.S Federal
Geographic Data Committee. To reduce the number of land cover classes, to better suit the needs
of this project, a re-classification was performed and land covers were aggregated based on leaf
type (i.e. broadleaved, needleleaved, grassland, wetland), leaf phylogeny (evergreen vs.
deciduous) and climate (temperate vs. tropical).
2.2.4 Soils
To negate the influence of soil mineral silica in dissolved silica concentrations found in rivers,
catchments with similar soils were compared. To geographically distinguish soil type regions,
the United States Department of Agriculture’s Natural Resources Conservation Service Soil
Survey Geographic (SSURGO) map was used. This data is based on a re-classification of the
FAO United Nations Educational, Scientific and Cultural Organization’s (UNESCO) Soil Map
of the World. The SSURGO map combined with a soil climate map, expressing 12 soil orders
according to Soil Taxonomy at three scales.
Page 31
18
2.2.5. Precipitation and NPP Data
Precipitation data used to reconstruct leaching rates was obtained from the Advanced Hydrologic
Prediction Service (AHPS) database through the National Oceanic and Atmospheric
Administration (NOAA). Files were extracted as monthly observed shapefiles for years dating
2005 to 2012 and imported into ArcGIS. Precipitation was averaged for drainages corresponding
to the selected twenty-six gauges. The data itself was measured as a 24-hour total summed per
month and is displayed as a grid of points with a spatial resolution of 4 x 4 km.
Net Primary Productivity data was obtained from images produced by NASA’s Earth
Observatory Team using TERRA/MODIS satellite imagery. Monthly values in g C m-2
day-1
were obtained for the twenty-six selected catchments using ArcGIS for years 2005 to 2012. For
catchments expressing DSi data as average annual values, NPP was summed to produce annual
averages. Net primary productivity for catchments that were used to display monthly variation
was shown as a daily sum. Net Primary Productivity satellite imagery for the United States was
used at a 1 km resolution.
Page 32
19
Table 2. Modeled Si fluxes, pools and rate constants for terrestrial systems with respective
calculations and variables. LF is the leaching factor, DF is the dissolution factor, NPP is net
primary productivity, %BSi is percent biogenic silica as net dry weight, SiD is the average
annual Dsi, F is the soil water flow, mp is phytolith mass, SSA is the phytolith specific
surface area.
Flux Calculation
Silica Flux LF · DF (NPP · % BSi)
Burial Production
BSi Storage Production – Dissolution
Dissolution (SiD · Q) / (mp · SSA)
DBSi Storage Dissolution - Leaching
LC Precip Vs. DSi r2
DF Dissolution / Production
LF (Silica Flux / Dissolution) · LC
Table 3. Phytolith specific surface area (SSA) and mass used for grasslands, wetlands,
coniferous and deciduous forests.
Phytoliths
SSA
(cm2/g) Mass (g)
Horsetail 928000 0.5
Larch 1950000 0.25
Elm 1210000 0.3
Page 33
20
Figure 2. Schematic of terrestrial biogenic silica model. Boxes in blue reflect dissolved
silica, boxes in white reflect silica in solid state. Dotted boxes refer to dominating processes
which influence the fluxes in the direction of the arrows. SAA refers to specific surface
area.
Page 34
21
Figure 3. Map of the United Sates of American showing point locations of the twenty six
gauges studied and corresponding land cover.
Page 35
22
Chapter 3
3.0 Model and Result Validation
Variables used for the construction of the TBSi cycle were approximated using various methods.
Magnitudes of these parameters from modern natural environments coincide with data
established as boundaries defining land cover types. The use of information derived from
existing systems allows for a unique model, nicely rooted by physical data as opposed to
theoretical. Consequently, justification for the selection of average parametric values correctly
describing a land cover is required and given through support from literature. In addition,
although environmental mass balances of Si are scarce, the fluxes calculated in this study
corroborate well with those determined by other researchers.
3.1 Watershed Characteristics
3.1.1 Watershed Productivity
Annual NPP values, used to estimate the quantity of biogenic silica produced, differ between the
four land cover regions (Table 4). By far wetland drainages expressed the highest NPP,
averaging 6978 ± 453 kg ha-1
y-1
. Although large, this value is supported by productivity studies
of marshes and wetlands which have reported among the highest production rates for terrestrial
ecosystems (Wieder and Lang, 1983; Rocha and Goulden, 2008). Such inflated NPP values are
believed to be attributed to wetland high carbon use efficiency (Lorenzen et al., 2001; Van Iersel,
2003). Second greatest NPP was expressed by grasslands, subsequently deciduous forests and
coniferous forests at 2823 ± 131 kg ha-1
y-1
, 2454 ± 105 kg ha-1
y-1
, and 1404 ± 394. kg ha-1
y-1
,
respectively. Within literature, grassland NPP is seen to fluctuate greatly, ranging from 940 to
4200 kg ha-1
y-1
(Hicke et al., 2002; Scurlock et al., 2002; Blecker et al., 2006). This variation is
greatly influenced by precipitation and temperature, as expected considering the effect these
factors have on the success of grasses (Blecker et al., 2006). This study’s used NPP for
deciduous forests tends to fall on the low side when compared to other studies (Norby et al.,
2002; Milesi et al., 2003), but is still comparable. Calculated coniferous forest NPP is also sound
Page 36
23
as it is in agreement with values ranging 1000 to 3000 kg ha-1
y-1
, determined for similar
vegetation (Gholz, 1982).
Biogenic silica content also differed among vegetation types found within land cover categories.
Grasslands showed the largest BSi content of 2.30 ± 0.13% of net dry weight. This value is
expected as Poaceae grasses have been found to contain the highest relative shoot Si
concentrations among forty-four other angiosperm clades (Hodson et al., 2005). Silica contents
of wetland vegetation, 1.91 ± 0.21% BSi, reflected values similar to that of other studies as well.
The mosses, horsetails, and aquatic grasses, which comprise this group, can have concentrations
of biogenic silica ranging from 2 to 28% BSi, globally (Schoelynck et al., 2009). These high
%BSi values for both grasslands and wetlands can be attributed to the lowered cost of
metabolizing silica. Conversely, biogenic silica of vegetation from both coniferous and
deciduous forests expressed low values at 0.84 ± 0.19, and 0.54 ± 0.11% BSi, respectively.
These values are also comparable with those from literature (Geis, 1978; Hodson and Sangster,
1999; Hodson et al., 2005).
3.1.2 Precipitation and Discharge
Calculated average annual precipitation from 2005 to 2012 was greatest for catchments
dominated by conifers, averaging 146.78 ± 13.74 cm (Table 5). Average precipitation found was
greater than others measured for coniferous forests in the western states. Precipitation data
obtained from field measurements suggests a range of 35.56 to 83.82 cm for this land cover type,
varying greatly on an annual basis and with local geography (Dodson and Root, 2013). Grassland
drainages expressed the lowest annual precipitation, averaging 62.89 ± 7.59 cm. This low value
is in agreement with biome measures made across the American Great Plains (Blecker et al.,
2006).
Average annual precipitation, discharge and catchment area was used to standardize DSi riverine
concentrations between watersheds. The relationship between discharge, precipitation and
drainage area, established for the twenty-six watersheds (Figure 4), shows a decrease in
discharge with a decrease in precipitation and drainage area. This relationship is expected, and is
a result of the positive linear relationship between drainage area and discharge (Menabde and
Page 37
24
Sivapalan, 2001) and precipitation and discharge (Knighton, 1998). Gathered data conforming to
this general trend allows for its use in predicting fluxes of DSi.
3.2 Watershed Silica Fluxes
3.2.1 Biogenic Silica Fixation Flux
Estimated biogenic silica production rates, based on NPP and %BSi content of foliage, greatly
vary among vegetation types but are also consistent with values estimated in literature (Table 6).
This study found that wetlands produced, on average 154.56 ± 28.83 kg Si ha-1
y-1
, certainly the
largest fixation rate among the four land cover types. Grasslands were found to produce less Si
than wetlands by half, approximately 65.65 ± 4.09 kg ha-1
y-1
. Coniferous and deciduous
catchments showed the lowest rates, 39.57 ± 13.60 kg ha-1
y-1
, and 44.33 ± 1.91 kg ha-1
y-1
,
respectively. However, because total biogenic silica from forests does not enter the soil system
annually, values were weighted for plant retention. This accounts for silica stored in stems and
other structures that do not constitute annual litterfall. True biogenic silica production of both
coniferous and deciduous catchments is 11.87 ± 4.08 kg ha-1
y-1
, and 13.29 ± 0.57 kg ha-1
y-1
,
respectively. Other studies that independently measured production of these vegetation classes,
predicted fixation rates to be greater than what we see with this model but are still within the
same order of magnitude (Table 6). Variation evident between production rates among studies
can be attributed to annual differences in net primary productivity used for rate estimations.
3.2.2 Biogenic Silica Dissolution Flux
Following BSi production, silica within the terrestrial cycle is subject to the process of
dissolution. Biogenic soil silica dissolution for both wetland and conifer dominated drainages
show the highest rates (Table 7). Coniferous silica exhibits a dissolution rate of 30.46 ± 11.62 kg
Si ha-1
y-1
, and wetland silica, 18.07 ± 5.13 kg Si ha-1
y-1
. Silica originating from grasslands has
the lowest dissolution rate, 4.7 ± 0.59 kg ha-1
y-1
. Grassland silica expressing the lowest
dissolution rate is unexpected considering the following: (1) rates calculated by other studies, (2)
high BSi production and (3) relatively large river DSi present in these catchments. Conversely,
conifer dominated regions expressing the highest dissolution was also unexpected. Low Si
production rates dictate dissolution should be low; however this finding is in agreement with
Page 38
25
rates determined by other studies. Unfortunately, little data is available concerning dissolution of
biogenic silica in wetland settings, casting uncertainty as to the accuracy of this models
prediction. Deciduous forest dominated catchments express a calculated average dissolution rate
of 8.14 ± 2.13 kg ha-1
y-1
, comparable with the rate determined by Bartoli (1983). The
unexplained differences in dissolution for land cover types can be attributed to several processes
and environmental parameters, ranging from soil aluminum (Al) substitution capacities to
phytolith size.
3.2.3 Biogenic Silica Storage Reservoir
Biogenic silica storage, which is calculated as the difference between BSi in a pool and the BSi
leaving that pool, can be divided into two storage compartments. One portion of this model’s
silica storage reflects Si that does not undergo transformation, a consequence of low dissolution
rates. Wetlands have the greatest storage of solid biogenic silica, 136.49 ± 33.90 kg ha-1
y-1
,
likely attributed to high wetland NPP, plant %BSi, and moderate dissolution. Grasslands and
deciduous forests reflect storage pools of 60.59 ± 3.96 kg ha-1
y-1
and 5.15 ± 1.97 kg ha-1
y-1
,
respectively. Conifer dominated forests displayed the lowest quantity of solid BSi in the soil
pool, -20.51 ± 11. 05 kg ha-1
y-1
, suggesting a system in which BSi is subject to a net loss.
The second recognized soil storage is that of dissolved biogenic silica (DBSi). This reservoir
consists of biogenic silica that has been subject to dissolution, but not leached. Conifer and
wetland dominated regions tend to express the largest DBSi storage, 25.76 ± 10.37 kg ha-1
y-1
and 17.81 ± 5.11 kg ha-1
y-1
respectively, grassland regions have the smallest pool, 3.12 ± 0.57 kg
Si ha-1
y-1
, while deciduous forests show intermediary storage, 7.76 ± 2.15 kg ha-1
y-1
. This
parameter is influenced by both dissolution and leaching rates. Dissolution tends to have a larger
influence than leaching attributed to the difference in magnitudes of both fluxes, as we shall see
shortly. The amount of dissolved silica stored in soils is proportional to the dissolution rate,
explaining the inflated storage of Si found in both conifer and wetland soils.
Both storage components sum to produce total BSi stored in soils. When relating modeled results
of all land cover types to literature (Table 8), coniferous storage shows the largest discrepancy.
While other sources suggest low storage of BSi in soils overlain by coniferous vegetation, they
Page 39
26
do not reflect values as low as the -20.51 ± 11. 05 kg ha-1
y-1
presented here. This amplified loss
can be likely accredited to the large dissolution flux of conifer phytoliths.
3.3 Biogenic Silica Riverine Flux
3.3.1 Leaching
Acting upon the dissolved biogenic silica pool in soils, are forces that ultimately cause Si
movement through the soil profile into topographic lows, such as rivers. The medium through
which DBSi moves is precipitation. Ideally, this relationship is a function of soil porosity,
percolation and soil type. The leaching factor in this study was calculated to reflect the
relationship between precipitation and dissolved lithogenic + biogenic silica flux, as constants of
proportionality (Figure 5). Three of the land cover types showed a fairly strong positive linear
relationship between precipitation and DSi fluxes. Conifer dominated regions express the largest
leaching factor, 0.78, proposing that 78% of silica found within the dissolved silica flux could be
explained by precipitation. Grasslands express a leaching factor of 0.57 and deciduous forests of
0.33. Wetlands, on the other hand, proved to have a very weak relationship between the two
variables, r2 = 0.25. In wetlands it is evident that precipitation may not be directly involved with
the amount of Si that is leaving those environments.
3.3.2 Riverine DBSi Estimation
Calculated silica fluxes include both biogenic and inorganic sources. To extract the biogenic Si
component, the effect of BSi leaching was considered on the USGS DSi fluxes for each land
cover type. Generally, the dissolved riverine Si fluxes represent a fraction of the dissolved Si
reservoir found in soils. Conifer dominated catchments showed to have the largest biogenic silica
flux, 4.70 ± 1.54 kg ha-1
y-1
, approximately a fifth of the DBSi soil pool. Grassland dominated
regions have the next largest biogenic silica flux of 0.903 ± 0.320 kg ha-1
y-1
. Deciduous and
wetland regions have the lowest biogenic silica fluxes, 0.384 ± 0.032 kg ha-1
y-1
and 0.258 ±
0.064 kg ha-1
y-1
, respectively, suggesting considerable Si retention within soils or other pools.
When comparing DBSi results of this study to other literature some disagreement is apparent
(Table 9). Estimated wetland and deciduous DBSi shows very low values while studies reflect
Page 40
27
those rivalling that of coniferous catchments. Although other estimates of wetland and deciduous
regions suggest higher DBSi values, these studies are few and may not be representative of
whole ecosystems. A literature review of DBSi found in waters of coniferous catchments
expresses the largest flux. This is in agreement with this study`s findings, and the reasons behind
this are speculated to be rooted in soil processes, described shortly. Also supporting this study`s
results, literature shows that grassland catchments reflect a DBSi flux ranging from 0.2-11 kg
ha-1
y-1
. This range correlates well with this study’s estimated grassland DBSi flux. Both
grassland and coniferous catchments show variation spanning two orders of magnitude. This
variation can be attributed to several factors relating to both biotic and abiotic processes.
3.3.3 Riverine DBSi Seasonal Variation
As an alternative measure to using leaching factors, dissolved biogenic silica was estimated by
separation of USGS DSi data into two monthly categories. One group consisted of DSi data from
May to September, known as the BSi reduction term, during which lithogenic silica is the
dominate component. The other group, known as the BSi accumulation term occurs from
October to April, and reflects a period during which the dominant component is biogenically
derived. This relationship is further supported by Ge/Si ratios and 30Si (White et al., 2012). For
all land cover types this biogenic component consisted of approximately 65% of DSi (Table 10).
To support the use of leaching rate to estimate DBSi, results were compared with data of DSi
collected from October to April. A residual analysis revealed an r2 value of 0.985 between the
two data sets (Figure 6). Between the DBSi flux calculated using leaching and the DBSi flux
from October to April data, the average difference was much smaller, 0.077 kg ha-1
y-1
. This is in
reaction to the difference between the leaching DBSi flux and DSi flux calculated using annual
USGS data, 1.29 kg ha-1
y-1
. This suggests the use of leaching to be much more comparable in
estimating DBSi than using solely DSi.
3.3.4 Soil Influence
To negate the influence of soils on biogenic silica fluxes, differences between catchment soil
types and lithogenic silica flux were analyzed. Dissolved lithogenic silica was determined as the
difference between calculated DBSi and DSi and soil orders were assumed to retain homogenous
Page 41
28
Si concentrations. Grassland catchments were dominated either by mollisol or alfisol soil orders.
Mollisol catchments averaged lithogenic silica fluxes of 0.41 ± 0.083 kg ha-1
y-1
while alfisol
catchments averaged 0.27 ± 0.139 kg Si ha-1
y-1
. Overlap in ranges suggests that differences
between the two are not significant. Conifer drainages were seen to be dominated by either
alfisols or inceptisols. Alfisol catchments averaged 1.6 ± 0.278 kg Si ha-1
y-1
and inceptisols
catchments, 2.8 ± 0.9 kg Si ha-1
y-1
, suggesting no statistically significant difference between the
two soils types. Deciduous forests dominated regions also showed no significant difference
between two dominant soils types, inceptisols which reflected 0.644 ± 0.089 kg Si ha-1
y-1
and
spodsols, 0.53 ± 0.06 kg Si ha-1
y-1
. All soils for wetland catchments were spodsols and as a
result riverine dissolved silica was not subject to soil based bias.
3.4 General Trends in Riverine Biogenic Fluxes
Following equation (1) biogenic silica fluxes leaving each distinct ecological region can be
calculated (Table 11). Silica fluxes have been shown to differ between land cover types as silica
contents of rivers are dependent upon vegetation type. This study shows that land cover does in
fact influence riverine silica content through production and dissolution of Si. However the final
process of leaching is dependent upon abiotic conditions of precipitation and soil characteristics.
In general, conifer dominated forests have the largest DBSi flux. This is counterintuitive
considering the low %BSi evident in evergreen vegetation, but appears to be compensated by
increased dissolution and high leaching. Conversely, grasslands which have a high %BSi show a
relative low riverine flux, yet high biogenic silica production. Reduction in the riverine flux can
be attributed to low dissolution rates relative to production, and resultantly high storage. Wetland
catchments express the highest silica fixation of all land cover types, yet some of the lowest
riverine fluxes. For this land cover type, approximately 90% of the silica produced remains in
soils as solid phytoliths, of the 10% that dissolves ~98% remains in soil solution while 2% leaves
catchments. Finally, deciduous forests which have rather low biogenic fixation have comparable
riverine fluxes to grasslands and wetlands. This can be attributed to the evidently low dissolution
rates of deciduous phytoliths, yet moderate leaching coefficient. Each of these environments is
unique in how silica is cycled within, and differences in magnitudes of fluxes and storage can be
attributed to inherent affinities of vegetation to silica, as well as a soil and climate processes.
Page 42
29
Table 4. Net primary productivity (NPP) and percentage biogenic silica net dry weight
(%BSi) of total plant weight for four analyzed land cover types.
Land
Cover
ANPP
(kg/ha · yr) SE %BSi SE
Grasslands 2823 ± 131 2.301 ± 0.130
Wetlands 6978 ± 453 1.971 ± 0.211
Coniferous 1404 ± 394 0.844 ± 0.186
Deciduous 2454 ± 105 0.541 ± 0.110
Table 5. Average catchment parametric values used for the calculation of dissolved silica
(DSi) fluxes for the four land cover types analyzed.
Land
Cover
Average
Annual
Precip
(cm) SE
Average
Annual
Discharge
(f3/s) SE
Drainage Area
Range
(ha)
DSi
Flux
(kg/ha
· yr) SE
Grasslands 60.96 ± 7.59 57.15 ± 18.55 20 719 to 116 286 1.584 ± 0.38
Wetlands 112.69 ± 2.09 40.88 ± 23.52 510 to 72 517 1.033 ± 0.23
Coniferous 146.78 ± 47.26 163.41 ± 68.51 5 638 to 26 590 6.022 ± 1.97
Deciduous 112.97 ± 13.74 83.048 ± 19.49 1 388 to 35 999 1.166 ± 0.07
Page 43
30
Table 6. Annual biogenic silica fixation rates from literature for wetland, grassland,
coniferous and deciduous ecosystems.
Land Cover Fixation (kg/ha · yr) Location Reference
Wetlands 500 Belgium Struyf and Conley, 2009
430 Poland Opdekamp et al., 2012
700 Africa McCarthy et al., 1989
~200 Global Carey and Fulweiler, 2012
154.56 ± 28.83 United States This study
Grasslands 22-26 United States Blecker et al., 2006
55-58 United States Blecker et al., 2006
59-67 United States Blecker et al., 2006
127 United States Alexandre et al., 2010
67 United States Alexandre et al., 2010
~25 Global Carey and Fulweiler, 2012
70 United States Carnelli et al., 2011
65.65 ± 4.09 United States This study
Coniferous
forest
8 United States Bartoli, 1983
15.8 Netherlands Markewitz and Richter, 1998
10.8-32.3 United States Garvin, 2006
29 United States Cornelis et al., 2010
42.2 United States Cornelis et al., 2010
2.1 United States Cornelis et al., 2010
24 United States Carnelli et al., 2011
11.87 ± 4.08 United States This study
Deciduous
forest
26 United States Bartoli, 1983
~50 Global Carey and Fulweiler, 2012
19.3 United States Cornelis et al., 2010
17.8 United States Cornelis et al., 2010
13.29 ± 0.57 United States This study
Page 44
31
Table 7. Annual biogenic silica dissolution rates from literature for grassland, coniferous
and deciduous ecosystems.
Land Cover Dissolution (kg/ha · yr) Location Reference
Wetlands 18.07 ± 5.13 United States This study
Grasslands 43-57 United States Blecker et al., 2006
43-51 United States Blecker et al., 2006
16-17 United States Blecker et al., 2006
103 United States Alexandre et al., 2010
74 United States Alexandre et al., 2010
50 United States Alexandre et al., 2010
62 United States Alexandre et al., 2010
4.7 ± 0.59 United States This study
Coniferous
forest
4 United States Bartoli, 1983
10-29.9 United States Garvin, 2006
30.46 ± 11.62 United States This study
Deciduous
forest
22 United States Bartoli, 1983
8.14 ± 2.13 United States This study
Page 45
32
Table 8. Annual biogenic silica soil storage for wetland, grassland, coniferous and
deciduous ecosystems.
Land Cover Soil Storage (kg/ha · yr) Location Reference
Wetlands 200 Belgium Struyf and Conley, 2009
154 United States This study
Grasslands 10-16 United States Blecker et al., 2006
4-13 United States Blecker et al., 2006
6-9 United States Blecker et al., 2006
12-24 United States Alexandre et al., 2010
5 United States Alexandre et al., 2010
60.59 ± 3.96 United States This study
Coniferous
forest
1 United States Bartoli, 1983
11.9 Netherlands Markewitz and Richter, 1998
0.8-2.4 United States Garvin, 2006
27.9 United States Cornelis et al., 2010
41.3 United States Cornelis et al., 2010
-7.4 United States Cornelis et al., 2010
-20.51 ± 11. 05 United States This study
Deciduous
forest
0 United States Bartoli, 1983
13.3 United States Cornelis et al., 2010
11.1 United States Cornelis et al., 2010
5.15 ± 1.97 United States This study
Page 46
33
Table 9. Annual dissolved silica flux from wetland, grassland, coniferous and deciduous
ecosystems.
Land Cover Riverine Flux (kg/ha · yr) Location Reference
Wetlands 19 China Nguyet et al., 2012
0.258 ± 0.064 United States This Study
Grasslands 6.3-11 United States Blecker et al., 2006
0.3-1.7 United States Blecker et al., 2006
0.2-0.5 United States Blecker et al., 2006
2.88 United States Alexandre et al., 2010
0.903 ± 0.320 United States This Study
Coniferous
forest
26 United States Bartoli, 1983
17 Netherlands Markewitz and Richter, 1998
15 United States Garvin, 2006
1.1 United States Cornelis et al., 2010
0.7 United States Cornelis et al., 2010
9.4 United States Cornelis et al., 2010
4.70 ± 1.54 United States This Study
Deciduous
forest
0 United States Bartoli, 1983
6.0 United States Cornelis et al., 2010
6.7 United States Cornelis et al., 2010
0.384 ± 0.032 United States This Study
Page 47
34
Table 10. Annual dissolved biogenic silica (DBSi) fluxes estimated using seasonality and
calculated leaching rates, in relation to annual dissolved silica (DSi) flux for four
ecosystems.
Land Cover Gauge
DSi Silica
Flux DBSi Silica Flux
Leaching
Rate
Oct to Apr R2-value
kg/ha · yr kg/ha · yr kg/ha · yr
Grasslands 05451210 5.55 3.788 3.164
05451080 1.987 1.511 1.133
06306200 1.21 0.924 0.690
Wetlands 01022890 0.985 0.158 0.246
02310947 0.617 0.522 0.154
02299950 1.497 0.813 0.374
Coniferous Forests 11264500 5.11 3.001 3.985
10343500 9.26 7.18 7.222
05014300 6.035 2.5 4.707
09196500 1.592 0.958 1.241
14161500 30.071 24.27 23.455
Deciduous Forests 01349950 1.248 0.965 0.411
01362380 1.446 0.69 0.477
01545600 0.962 0.622 0.317
04063700 0.795 0.49 0.262
01364959 1.206 0.73 0.397
01422747 1.336 0.904 0.440
Page 48
35
Table 11. Modeled average catchment fluxes for each land cover type analyzed and corresponding
coefficients.
L
Land Cover
Grassland
(kg/ha · yr)
Wetland
(kg/ha · yr)
Coniferous forest
(kg/ha · yr)
Deciduous forest
(kg/ha · yr)
Burial 65.65 154.56 9.94 13.26
BSi Storage 60.95 136.49 -20.51 5.15
Dissolution 4.70 18.07 30.46 8.14
DBSi Storage 3.12 17.81 25.76 7.76
Leaching 0.903 0.258 4.70 0.385
Dissolved Riverine
BSi Flux 0.903 0.258 4.70 0.385
Dissolution Factor 0.071 0.117 3.06 0.612
Leaching Factor 0.19 0.014 0.15 0.047
Leaching constant 0.57 0.25 0.78 0.33
Figure 4. Depiction of relationship between precipitation, discharge and drainage for the
twenty-six gauges analyzed. Showing decrease in discharge with decrease in drainage area
and precipitation.
Page 49
36
Figure 5. Dissolved silica fluxes and annual precipitation relationship between four studied
land cover types depicting leaching coefficients (r2 values). A. Grasslands B. Wetlands C.
Coniferous forests D. Deciduous forests.
Predicted DBSi Flux (kg ha-1 y-1)
0 5 10 15 20 25
Aug to A
pr
DB
Si F
lux (
kg h
a-1
y-1
)
0
5
10
15
20
25
30
r ² = 0.985
Figure 6. The near 1:1 ratio of predicted dissolved biogenic silica flux using leaching
coefficients and seasonal segregation.
Page 50
37
Chapter 4
4.0 Discussion
4.1 Biogenic Si Contributions
At catchment scales, DSi flux is a function of the following: (1) geology, (2) hydrology, (3) soil
development, and (4) biological processes (phytolith formation). Conley (2002) estimated annual
fixation of phytolith silica at 2 to 6 Gt Si y-1
, most of which is added as litterfall to soil surfaces.
Although there is uncertainty regarding factors controlling the relative solubility of different
types of phytoliths, there is consensus that contributions of phytolith dissolution to DSi export
could be substantial. More recently, soil-plant systems have been detailed using geochemical
tracers, particularly Ge/Si ratios and 30Si. Fractionation between germanium (Ge) and silica can
be used to trace weathering of silica and dissolved silica from biogenic origins (Kurtz et al.,
2002; Derry et al., 2005). Ge/Si ratios are enriched in Ge for samples dominated by secondary
minerals, while biogenic silica polymerized by plants is depleted (Figure 7) (Delvigne et al.,
2009; Opfergelt et al., 2010; Cornelis et al., 2010). Alexandre et al., (1997) found that Si released
from phytolith dissolution is twice that of Si released from silicate weathering in tropical
systems. In these environments this is expected as depletion of mineral Si and high Si uptake
rates by biomass are evident (Lucas et al., 1993). Ge/Si ratios of stream waters from Hawaiian
basaltic catchments suggest biogenic contributions of up to 90% (Derry et al., 2005). Other
estimated contributions of phytolith dissolution are 30% in a coniferous Siberian forest
(Pokrovsky et al., 2005), 75% in a Congo rainforest (Alexandre et al., 1997) and 47 to 74% in a
California grassland (White et al., 2012). The biogenic silica contributions estimated for each
land cover type from this study ranges from 54% for humid wetlands to 60% for temperate
broadleaf deciduous forests. For a wetland catchment BSi contributions ranging from 15% to
80% emphasize the importance of influencing factors other than biological processes.
4.2 Conifer Anomaly?
Estimated mass balances of Si output in conifer dominated catchments reflect values much larger
than expected considering such low biogenic silica inputs. This is especially true when
Page 51
38
comparing coniferous catchments to grasslands, whose vegetation is comprised of Si
accumulating plants. This trend of high silica flux leaving coniferous catchments has been
recorded in several independent studies, but not addressed to any great extent (Garvin, 2006;
Cornelis et al., 2010).
The relationship between silica uptake (production) and dissolved outputs has been found to be
negatively correlated (Figure 8). To explain this, it has been suggested that DSi output exceeds
Si production when DSi released by mineral dissolution does not contribute to the BSi pool of
vegetation (Cornelis et al., 2010). In cases where this relationship still holds but fixation is less
than output, vast quantities of Si on an annual basis must be retained in soils or vegetation. The
fact that BSi pools of vegetation act as a sink has been largely recognized in other studies of
forested and grassland ecosystems (Lucas et al., 1993; Aexandre et al., 1997; Giesler et al., 2000;
Gerard et al., 2002). The size of this pool, however, varies between vegetation as seen
previously, where conifer production is the lowest among the most common land cover types. In
addition to Si retention within vegetation, low Si flux rates would suggest large Si storage within
soils. This is indeed found to be the case for grasslands and not for coniferous regions (Sommer
et al., 2000; Conley, 2002, Melzer et al., 2011). If vegetation is rapidly recycling nutrients,
inputting silica into soils, it is likely that primary silicate weathering, and soil DSi, will be
decreased (Kelly et al., 1998). However, Si uptake by vegetation directly impacts soil formation
through dissolution, increasing primary and secondary silicate mineral formation (Lucas, 2001).
These two processes both promote and supress the availability of silica in soils, and may explain
why in grasslands we see high silica uptake with high solid silica soil content. So the question
becomes, why do coniferous catchments export nearly all of their fixed Si, while grasslands
export only a fraction? To answer this question process affecting dissolution and leaching must
be addressed.
4.3 Phytolith Dissolution
Early experiments performed on the dissolution of phytoliths showed that forest BSi is
approximately 10 to 15 times more soluble than grasses (Wilding and Drees, 1974). In addition,
conifer phytoliths have been found to be least stable when compared to those of grassland and
Page 52
39
broadleaf vegetation (Bartoli and Wilding, 1980). Three main factors have been implicated in
drastically altering BSi dissolution: (1) phytolith surface area, (2) presence and abundance of Al
in soil and plant tissues, and (3) soil pH.
4.3.1 Surface Area Size and Dissolution
Specific surface areas (SSAs) of phytoliths have been proposed to influence physical processes
both in plants and in soils (Bartoli, 1985; Piperno, 2006; Li et al., 2013a; 2013
b). Phytolith
dissolution rates increase substantially with greater surface areas by increasing solubility
(Fraysse et al., 2006; 2009). This occurs since increasing surface area leads to an increase in non-
proton/hydroxyl reactive surfaces, allowing for increased deprotonation of surface silanol bonds
(Fraysse et al., 2009). In a phytolith dissolution study carried out by Fraysse et al., 2006,
dissolution of bamboo phytoliths revealed an increase of two orders of magnitude for specific
surface areas of 5.18 m2/g to 159 m
2/g. In addition to phytoliths, the dissolution rates of
diatomaceous frustules also increase with specific surface area (Van Cappellen et al., 2002;
Loucaides, 2010). Species specific variation in surface areas has been found to cause differences
in dissolution efficiency of several orders of magnitude (Martin-Jezequel et al., 2000; Dixit et al.,
2001; Ryves et al., 2001). Conifer phytolith specific surface areas average 195 m2/g, which are
significantly greater than those of grasses, 92.8 m2/g, and deciduous trees, 121 m
2/g (Fraysse et
al, 2009). Large specific surface areas found for conifers, and related increase in solubility may
explain why large dissolution rates are evident within coniferous catchments.
4.3.2 Aluminum Induced Reduction in Dissolution
Aluminum ions are toxic to plants and are repressed from plant tissues by the presence of an
endodermis. However, this root layer is not completely effective as Al can often be detected in
shoots and leaves of some plant species (Hodson and Sangster, 1999). Aluminum concentrations
found in grasses are low, while conifer species have been found to accumulate far more Al into
plant tissues (Carnelli et al., 2001). This is the inverse with Si concentrations found in respective
vegetation types. This inverse relationship between Si and Al is expected since silica has been
found to mitigate Al toxicity (Hodson and Evans, 1995; Cocker et al., 1998). Aluminum in
coniferous species has been found in concentrations of up to 28.3 mol g-1
dry weight, while
Page 53
40
cereals have been found to contain concentrations of no more than 5 mol g dwt-1
(Hodson and
Sangster, 1999). The presence of aluminum and its adsorption onto BSi in plant tissues and soils
has been found to deter phytolith dissolution (Dixit et al., 2001; Rickert et al., 2002). Adsorption
of Al by siliceous particles results in co-deposition of Si-Al insoluble aluminosilicates, reducing
dissolved silica mobilization (Cappellen and Qiu, 1997).
The effect aluminum has on silica dissolution from both organic and inorganic origins and the
knowledge that conifers tend to show higher concentrations of Al, would suggest conifers have
low dissolution rates. However, this is contradictory to what is seen in both this study and other
literature. A study performed by Cornelis et al., 2010 shows that black pine takes up and deposits
considerable amounts of Al, 3.3 kg ha-1
y-1
and 12.2 kg ha-1
y-1
, respectively. Yet, this land cover
still reflects a Si soil water output of 9.4 kg ha-1
y-1
. A Deciduous tree, European beech, which
takes up and deposits only 0.8 kg Al ha-1
y-1
and 5.9 kg Al ha-1
y-1
, respectively, still only releases
6 kg Si ha-1
y-1
(Cornelis et al., 2010). To assess this relationship more fully, the magnitude of Al
influence on each respective environment is needed. In this case, it is entirely likely that although
aluminum induced stabilization of Si reduces DBSi export, it is muffled by the surge of
dissolution caused by higher SSA of coniferous phytoliths. Conversely, it can be argued that
dissolution of phytolith BSi is severely halted by Al adsorption in the surface soil layers where
phytoliths and Al (Nikodem et al., 2007) are in abundance. The progressively high DSi export
could then be mostly attributed to mobilized lithogenic Si, provided that BSi contributions for
coniferous catchments have been over projected.
4.3.3 Influence of Soil Acidity (pH)
Dissolution kinetics of BSi, in both terrestrial and marine environments increases with extremes
of pH. Highly alkaline and acidic soils experience increased deprotonation of silanol groups,
which form biogenic silica molecules. This process facilitates the breakage of siloxane bonds,
believed to be the rate-limiting step in the dissolution process of silica (Dove and Elston, 1992).
The effects of pH on the kinetics of silica dissolution have been established for quartz (Dove and
Elston, 1992) and BSi (Fraysse et al., 2006; 2009; Loucaides et al., 2008). In a recent study,
Loucaides et al. (2008) demonstrated that dissolution between pH of 6.3 and 8.1 double in rate;
Page 54
41
in a study by Fraysse et al. (2008) dissolution was seen to increase by a factor of 15 for every
order of pH from 4 to 12, as well as for pH solutions from 3 to 1 (Figure 1). This would suggest
that catchments containing very acidic soils, like coniferous podzols, and those containing very
alkaline soils, like grassland chernozems, should express the highest dissolution rates.
The relationship between soil pH and calculated dissolution rates holds weakly for this study’s
reviewed land covers. Coniferous podzols show pH values ranging from 3 to 4 (Sommer et al.,
2006; Neubauer et al., 2013), while grassland soils express pH values ranging from 5.5 to 8.4
(Saccone et al., 2007). Using the dissolution and pH relationship developed by Frassye et al.
(2009), dissolution in grasslands should be nearly a factor of 10 greater than in coniferous
forests. This does not appear to be the case; in fact the dissolution rate of grassland silica is 6
time less than that of coniferous catchments. The calculated conifer silica dissolution rate has
large standard errors, reflective of great variation in DSi concentrations of river waters. This
variation may arise due to differing lithology, topography, or a yet unexplored confounding
factor. The dissolution calculation requires the concentration of dissolved Si in soils that is
leaving the system. This value is scarce in literature and often not measured, and was thus
approximated using DSi concentrations found in riverine waters, suggesting an alternative source
of error.
4.4 Soil Silica Transport to Streams
In tandem with silica dissolution, the rate of leaching also influences silica export. The
availability of silica along a soil profile influences riverine Si deposition, contingent upon
whether dominant water movement process are occurring at the surface as runoff, or as
horizontal flow at depth. If deposition and dissolution of Si is greatest in organic and A horizons
with water translocation occurring in soils closer to the surface, then biogenic silica contribution
to rivers should be great. Conversely, if dissolution is occurring lower down the soil profile, and
water translocation is occurring at the surface, Si export will be low (White et al., 2012).
However, if in this case the dominant form of water movement is subsurface lateral flow, then
lithogenic contributions to silica should be greater than biogenic, as phytolith deposition is
Page 55
42
limited to the surface. Transportation of silica laterally to rivers is thus a function of dissolution
location, and water movement processes defined by precipitation.
For nearly all systems phytolith deposition and dissolution will appear greatest near the surface
in the soil A horizons (Blecker et al., 2008; Fishkis et al., 2010; White et al., 2012). Following
episodes of rainfall, water percolation and gravity will transport this dissolved silica to regions of
low water potential, such as rivers or other bodies of water. In the case of grasslands, clay to silty
clay loam soil textural classes have low hydraulic conductivities of < 0.13 to 2 cm hr-1
(O’Green,
2012). Low hydraulic conductivity decreases infiltration rate and percolation lowering the rate of
water outputs, and dissolved silica within. This is especially true under conditions of low
precipitation. In cases where precipitation surpasses infiltration, runoff may not have opportunity
to absorb dissolved silica and direct it towards rivers. Cases such as these are evident in
grassland environments, dominated by silty clay textured soils which have low infiltration rates
on account of lowered pore connectivity and colloid formation (Kurz et al., 2005). Coniferous
podzol silty loams have hydraulic conductivities of 2 to 12.7 cm hr-1
(O’Green, 2012). Increased
connectivity and pore space in forest soils, as well as higher annual precipitation suggests the
potential for greater nutrient leaching rates than in grasslands. The stratigraphic location of silica
dissolution occurs near the surface of both grasslands and coniferous forests; however, export of
this silica is limited by the rate of water percolation/infiltration through soils, substantially lower
for grassland ecosystems.
4.5 Wetland Silica Retention
Wetland systems are the link between terrestrial and marine environments, and as a result
interact strongly with river biogeochemistry. Modeled wetland systems show high biogenic silica
production moderate dissolution, but low leaching rates resulting in a great DBSi storage
component. Wetland soils contain between 0.6 and 0.9% solid biogenic silica by weight,
considerably lower than other soils (Norris and Hackney, 1999; Struyf et al., 2005). Low solid Si
concentration is consistent with high dissolution rates, and further supported by large quantities
of dissolved silica (Struyf et al., 2007). Although large amounts of silica are being fixed and
dissolved in these environments, we do not see increased export, or at least a proportional
Page 56
43
relationship between rates of land cover classes analyzed. In a study on the effectiveness of
wetlands at retaining nitrogen and phosphorous, 80% of wetlands held onto 70% of nitrogen and
60% of phosphorous inputs (Fisher and Acreman, 2004). A DSi budget constructed by Struyf
and Conley (2009), shows that wetlands are capable of preserving up to 21% of Si dissolved in
soils.
Nutrient cycling within wetlands is a function of catchment features (drainage shape, water
discharge, vegetation, and hydrology) which manipulate the length of time it requires for water
to pass. Low discharge of wetland waters leads to increased residence of nutrients (van der Valk
and Arnold, 2009). Consequently, low discharge seen in the three wetland catchments may
explain why silica export and recycling proved to be low. In addition to water flow,
sedimentation can also act to sequester biogenic silica within soils. The process of sedimentation
contributes to wetland functioning, attributed to sediment sorption of nutrients and other
minerals (Craft and Casey, 2000). Sedimentation rates in wetlands range on the order of 271 to
712 g m-2
d-1
which greatly acts to withdraw nutrients from the water column (Nahlik and
Mitsch, 2008).
The difference in DBSi and the silica dissolution flux used to estimate the storage DBSi in soils
can also be greatly influenced by diatoms. Fresh water diatom metabolises of silica depletes
dissolved riverine concentrations (Martin-Jezequel et al., 2000; Struyf and Conley 2009;
Carbonnel et al., 2009; Luu et al., 2012; Viaroli et al., 2013). As a result DSi concentrations may
be biased through the amount of Si recycled by these diamateous organisms. This is especially
true if diatom abundances differ between wetland catchments, and separation of DSi into
seasonal components to extract DBSi did not yield reasonable estimates of the biogenic portion.
4.6 Regional Implications
The TBSi simulations from this study have many implications for the functioning of Si within
the terrestrial biosphere as we have seen, and globally. At a regional scale, biogenic silica input
within conifer dominated environments is low. However, quick dissolution and increased
mobility of silica through soils in these environments allows for a large silica flux. In these
coniferous systems, the solid biogenic silica pool in soils is very low or negative suggesting that
Page 57
44
the soil storage-export balance is not in steady state. Either disequilibrium is occurring or nearly
all biogenic silica produced is being dissolved. In grassland systems the moderate annual BSi
production constituent undergoes weak dissolution. This dissolution rate is a consequence of low
phytolith surface area and soil hydraulic conductivity. As a result of such low dissolution these
environments accumulate large quantities of solid biogenic silica in soils. Low leaching rates in
grasslands, attributed to a weak correlation between precipitation/riverine Si flux and low
hydraulic conductivity of soils, dictates that these environments have low DBSi outputs.
Wetlands which reflect a high Si production flux and dissolution rate show a surprisingly low
riverine flux. This inconsistency is likely a result of nutrient retention within soils. In this case
retention is a function of low discharge and sedimentation. Deciduous catchments appear to be a
sensible case in both flux and storage of silica. Moderately low annual production, storage and
dissolution show a proportional DBSi flux in-between that of grassland and wetland
environments. When interpreting the results of this study at a broader scale, it becomes apparent
that coniferous catchment Si export dominates, followed by grasslands, deciduous forests and
finally, wetlands.
4.7 Vegetation Influence
The influence of land cover class on dissolved biogenic silica concentrations is dominated by
process of Si production and dissolution. Coniferous catchments which show a large riverine
flux, attributed to BSi dissolution, is a function of vegetation type owing to phytolith specific
surface area. For land covers such as grasslands and wetlands, high biogenic Si production and
low riverine BSi flux, can be explained by biologically influenced processes of dissolution as
well. However in these systems having low soil porosity or high retention, are abiotic processes
influencing leaching, which also lead to reduced DBSi export. In the case of coniferous and
deciduous forests, riverine BSi flux appears to be more greatly influenced by vegetation type
than grassland and wetland environments. Forested regions appear to have a higher biological
control on BSi leaving these environments. This is not to say that there is no relationship
between vegetation cover and riverine flux for grasslands and wetlands. This relationship is
simply reflected through abiotic characteristics associated with their respective vegetation types.
Page 58
45
4.8 Sources of Error
Within in this study several assumptions were made concerning both data collection and
calculation. Data collection was severely limited by data available temporally and spatially. This
concept of Si influencing the terrestrial biosphere is relatively new, and the collection of data
used to define this relationship, has just recently been implemented. The use of only twenty-six
catchments begs the question of whether or not the depiction of land cover types is
representative. For grasslands, wetlands and deciduous forests low standard errors suggest that
land covers are depicted to a fair degree of accuracy. For coniferous catchments, which reflect
large standard errors, an explanation needs to be invoked. Large differences in conifer flux
calculations are a result of increased variation in dissolved silica concentrations measured in
waters. These variations may be accounted for by differing lithology, local soil type, coniferous
taxa (which are seen to have great variation in Si content between species), topography, or yet
another factor.
Amalgamation of direct measurements for the creation of generalizations, specifically for use in
a mathematical relationship, often leads to errors which must be recognized. Inaccuracies of data
collection could lead to misinterpretation of results and overall amplification of error. This is
particularly significant if a series of equations are nested and rely on previously calculated
values, as is the case in this study. In the case of wetland NPP estimations, it is entirely possible
that values are exaggerated. Particularly in wetland regions, photosynthesis and carbon
production can be greatly influenced by NPP contributions from algae, and other phytoplankton.
As these organisms do not produce Si to any great extant, Si production for this land cover type
could be overstated. For dissolution calculations, riverine DSi was assumed to be a suitable
representative of dissolved silica in soils. Dissolution in soils can be greatly influenced by soil
specific process, such as bacterial Al-silicate formation (Konhauser and Urrutia, 1999). Riverine
DSi may not be directly related to soil DSi on account of unique processes, but once again data
availability dictated the information used within the model.
In addition to this, DSi concentrations measured in rivers by USGS contractors does not
discriminate against Si biomineralized by fresh water diatoms. Riverine cycling of Si is similar
Page 59
46
to oceanic, especially in how it is influenced by diatoms. Diatoms take up dissolved silica;
convert it to biogenic forms for skeletal formation, and upon death biogenic silica is mobilized.
This means that inorganic silica is converted to biogenic in river water by diatoms, and not solely
from vegetation as is assumed in this study. Diatom influence on biogenic silica can influence
the biogenic component extracted from dissolved values through inflation. This study was
limited by available data and as more research is performed on these systems, adjustments and
corrections should be implemented accordingly.
Page 60
47
Figure 7. A. Depiction of the relationship between Ge/Si ratios during the growing season
(Green), and die back (Brown). B. Relationships of 30
Si/ 28
Si isotopes during the growing
and winder seasons. After White et al., 2012.
Si Fixation (kg ha-1 y-1)
0 20 40 60 80 100 120 140
Si R
ive
rine
Flu
x (
kg h
a-1
y-1
)
0
2
4
6
8
Deciduous Broadleaf
Coniferous
Grassland
Wetland
Figure 8. Relationship between annual silica production and export. Coniferous regions
show low fixation but high export, while grasslands show the reverse.
Page 61
48
Chapter 5
5.0 Case Study
5.1 Global Oceanic Biogenic Silica Input
The effects of the TBSi cycle previously established can be applied globally to assess the impact
of vegetation on oceanic inputs. It has been seen that vegetation type can influence the amount of
silica entering riverine settings through silica production and dissolution. Recognizing how land
cover can influence global dissolved oceanic silica concentrations will lead to a better
understanding of the global silica cycle. The connection between oceanic Si and diatoms is well
established within literature and of significant importance since diatoms are responsible for
nearly half of oceanic productivity (Nelson et al., 1995). Diatoms are first seen to arise during
the Jurassic and reflect rapid radiation during the last 40 million years. Diatom influence on
oceanic primary productivity has led to great interest in clarifying factors that led to this
diversification. One current proposition is that geochemical coupling between terrestrial
grasslands and marine ecosystems through global silicon cycles stimulated this radiation
(Falkowski et al., 2004; Conley, 2002).
Diatoms have an absolute requirement for silica of 2 µmol (Egge and Aksnes, 1992), and as
such, their prominence in modern oceans is thought to have been promoted by processes that
increased the flux of silica into marine systems (Egge and Aksnes, 1992). At concentrations of
less than 2 µmol diatoms were seen in low abundance, a result of nutrient starvation. The rise of
grasslands in the Miocene is debated as being a trigger leading to increased weathering and Si
biomineralization. This theory is supported though synchronous diatom diversification
(Falkowski et al., 2004). Grasslands take up substantially more silica per unit carbon than many
other terrestrial systems. Through doing this, they create a substantial phytolith and silica pool
within their soils. Some of this silica is released during small scale events, such as fires (Pisaric,
2002), while the remaining silica is stored. Large scale tectonic pulse events, such as earth
quakes, would also substantially increase rates of Si release from grasslands regions. This
principle of grasslands influencing diatom radiation is reliant upon the assumption that great
Page 62
49
quantities of silica are more readily released from grasslands than other ground types. The
constructed biogenic silica cycle of this study suggests that this is not the case as we see conifer
dominated catchments having a greater biogenic Si flux. However, this is not to say that
grasslands did not influence diatom radiation as on a global scale ground cover area would
greatly influence oceanic silica concentrations. If grasslands dominate continental cover it is still
likely that biogenic Si exports from these systems was considerable. In addition to biogenic Si, if
grassland enhanced weathering increased fluxes of lithogenic silica, oceanic concentrations
would similarly increase.
To reiterate the objectives stated in the introduction, the aims of this case study were as
follows:
(1) Define land cover change on global oceanic dissolved silica, and
(2) Determine if changes in terrestrial ecology during the Oligocene triggered diatom
evolution/radiation
5.2 Methods
5.2.1 Oceanic Si Estimation
To assess the potential impact of biogenic silica fluxes and defining land cover classes on diatom
evolution, global terrestrial export values and oceanic silica concentrations were estimated.
Global terrestrial silica fluxes were projected for the Eocene, Oligocene, Miocene, and Pliocene,
and used to estimate oceanic BSi concentrations. These time periods were selected to represent
oceanic conditions prior to and following diatom diversification, which occurred during the
Eocene-Oligocene transition. These values were subsequently compared to the dominance
threshold of 2 µm as a diatom growth requirement, determined by Egge and Aksnes (1992) for
contemporaneous oceanic conditions. The global biogenic silica flux into oceans was calculated
using the equation;
∑ , (7)
Page 63
50
where is Area (ha) and is the total amount of biogenic silicon lost (kg ha-1
y-1
) for each
land cover class. Tropical and temperate forest silica fluxes were approximated using the average
deciduous DBSi flux, while grasslands and coniferous forests were assigned their respective flux
values calculated using the TBSi model. Using the calculated oceanic flux, oceanic silica
concentrations were estimated with;
Tsi = / , (8)
where is the flux of terrestrial silicon to oceans (kg) and is volume of ocean water (l).
Concentrations were calculated for total oceanic volume, 1.3 billion km3 of water, and a top layer
of 5 m. Total ocean volume was obtained from Garrison (2005). A top layer of 5 m was selected
representing ocean diatom and dissolved Si distributions.
5.2.2 Palaeo-Land Cover Distribution
To calculate terrestrial silica fluxes across the Cenozoic, reconstructed global vegetation
distributions were collected and input into ArcGIS to obtain areal extents in km2
(Figure 9 to 12).
Eocene and Oligocene distributions were obtained from Fine and Ree (2006) and were created
using modern day correlates, composite parameters and global carbon models (GCMs). The
Miocene data was adapted from Pound et al. (2011), which defined boundaries using TEVIS
(Tertiary Environments Vegetation Information System), palaeogeography, orography and sea
surface temperature. Finally, the Pliocene extents were obtained from Dowsett et al. (1999) and
were estimated using the Pliocene research interpretation and synoptic mapping 2 (PRISM2)
dataset.
5.3 Results and Discussion
5.3.1 Eocene to Pliocene Land Cover Change
Land cover distributions from the Eocene onward to the Pliocene were reconstructed using
various proxies. Proxy accuracy increases as more geologic and fossil data becomes available,
thus maps recreated from the Miocene to Pliocene are likely more accurate. Interpretation of
Eocene-Oligocene maps must be done so using caution knowing that reconstruction is based off
discontinuous data. During the Eocene, vegetation largely consisted of either temperate or
Page 64
51
tropical forests (Figure 9) (Ziegler et al., 2003; Fine and Ree, 2013) with some dispersed
grasslands (Stromberg, 2011). The Oligocene is characterized by the expansion of grasslands and
formation of grassland cover types that thrived near the tropics (Figure 10) (Lunt et al., 2007;
Stromberg, 2011; Fine and Ree, 2013). C4 grassland expansion during the Miocene dominates in
the topics displacing tropical forests (Pound et al., 2011; Stromberg, 2011) while in the nothern
latitudes coniferous forests begin to take hold as a prominent land cover type (Figure 11).
Coniferous fossil evidence suggests evolution during the Permian, however prominence of this
functional type did not take hold until after the Cretaceous, when frost tolerance developed
(Opalinska and Cowling, 2013). During the Pliocene, grasslands expanded from Africa to much
of southwest Asia and coniferous forests expanded further within the high latitudes (Figure 12)
(Haywood et al., 2004). We also see development of desert and Arctic tundra cover types during
this time. As a general progression from the Eocene to Pliocene we see a change from temperate
and tropical forests to grassland and coniferous dominance (Table 11).
5.3.2 Eocene to Pliocene Oceanic Si Change
Changes in land cover distribution and dominance from the Eocene to the Pliocene will affect
global BSi fluxes (Table 12). Following the Oligocene, global biogenic silica flux is seen to
double, doubtlessly influenced by the presence of vast grassland and coniferous forest expanses.
Miocene and Pliocene biogenic silica fluxes are dominated by coniferous forests where 85% of
the DBSi flux is attributed to conifers. Prior to the Miocene the dominant Si contributor was that
of temperate forests. Tropical rainforests appear to have the lowest impact factor on DBSi input
into oceans. The dissolved silica leaving the tropics was estimated using the deciduous DBSi
value. This likely under-exaggerated BSi leaving these systems as the tropics are subject to
differing precipitation and soil conditions (Cornelis et al., 2011). Modern BSi export values
estimated for the Amazonian rainforest are several times larger than that of temperate deciduous
forests, and rival that of grasslands (Lucas et al., 1993).
The increased flux of DBSi in the Miocene and Pliocene, attributed to grasslands and conifers
forests, reflects an increase in oceanic Si concentrations. Following the expansion of these two
functional types, oceanic concentrations of biogenic silica are seen to double. Oceanic
Page 65
52
concentrations were predicted for the whole ocean, but it is likely that circulation of silica would
lead to increased concentrations nearer the surface. Predicting oceanic concentrations for the top
5 m, where modern diatom abundances are greatest, suggests values of 0.069, 0.054, 0.019, and
0.028 µmol l-1
from the Pliocene, Miocene, Oligocene and Eocene, respectively. Modern oceanic
BSi flux was calculated with current distribution of land cover types and revealed a value of 0.42
Tmol Si year-1
. This value is less than the biogenic silica estimate of 1.1 Tmol Si year-1
devised
by Treguer and De La Rocha (2012). However, it is likely that 0.42 Tmol Si year-1
is understated
as it does not take into consideration inputs from groundwater or windblown sources.
5.3.3 What this means for Diatom Radiation
Diatoms have a silica requirement that must be fulfilled in order to allow polymerization of
skeletal structures. This requirement dictates that water must contain concentrations of at least 2
µmol of Si. In order for biogenic silica to influence diatom growth, diatoms must either
metabolize biogenic silica at a faster/more efficient rate, or the input of biogenic silica must be
substantial in magnitude. Capellacci et al. (2012) showed that for two diatom species, growth
rates were highest in the presence of dissolved silica derived from crystalline quartz. If
vegetation dynamics influenced diatom radiation, then it must have been through the change in
sheer magnitude of DBSi during the Miocene. Predicted oceanic concentrations exhibit values
that are far less than required, but do show a substantial increase with the presence of grassland
and coniferous forest expansion after the Oligocene. This increase in biogenic silica during the
Miocene could then be responsible for diversification of diatoms. These concentrations are solely
measured for biogenic silica, and the incorporation of lithogenic would undoubtedly increase
these values.
Modern flux studies have shown heterogeneity in global riverine silica inputs (Bernard et al.,
2011). Coastal regions near the Amazonian and Ganges river basins have the highest dissolved
Si outputs attributed to high discharge and drainage areas. It is not too impractical to assume that
in these regions biogenic silica fluxes would also be rather high and allow ocean waters to attain
values greater than 2 µmol of Si at least near continental shelves. Thus, if the increase in
biogenic silica during the Miocene influenced diatom diversification, it would have done so
Page 66
53
nearer river outlets on continental margins. While silica presence is the main contributing factor
for diatom persistence, other nutrients and their distribution would also influence diatom
radiation. For biogenic silica increase to have triggered diatom radiation, lithogenic contributions
must have remained close to equal through time. This study’s estimated oceanic concentrations
assume a homogenous distribution of silica which is not the case in modern oceans (Durr et al.,
2009; Bernard et al., 2011). Nonetheless total rise in oceanic silica may still have had an
influence on diatom radiation. To make more concrete inferences using average oceanic
concentrations, influence of the lithogenic component must be known.
5.3.4 What this means for Grasslands as an Instigator
The influence of grasslands as primary initiators leading to the diversification of diatoms during
the Miocene may not be as momentous as previously presumed. This is especially true when
analyzing grassland influence on biogenic silica contributions. Teasing apart grassland and
coniferous forest contributions for DBSi is challenging especially when performing analysis at a
global scale. A large catchment dominated by grasslands may elevate silica concentrations near
the point of export enough to promote diatom diversification regionally. However, this is more
likely to be the case with coniferous forests, which have a greater control on the quantity of
DBSi exported to oceans. The DBSi oceanic inputs were calculated at steady state on annual
bases. Large scale pulse events acting to release large quantities of silica stored in soils, would
certainly influence ocean concentrations. Following disturbances such as fires nutrients are
released quickly and can reflect losses of up to 85% (Laclau et al., 2002; Pisaric, 2002). Other
natural disturbances such as earthquakes and flooding could also potentially increase grassland
DSi export. Thus, grasslands may not have contributed greatly to the annual influx of BSi to
oceans, however grassland silica storage has potential to increase oceanic DSi concentrations
following disturbance.
5.3.5 Silica Retention
The retention of dissolved silica within fluvial systems has proved to decrease DSi exports to
coastal waters. Dissolved silica retention can occur due to anthropogenic factors such as
damming and eutrophication, and through wetland immobilization (Friedl et al., 2004; Humborg
Page 67
54
et al., 2008). The construction of dams have been seen to increase diatom growth due to
increased residence times, but ultimately leads to intensive sedimentation and net losses of silica
to reservoir sediments (Koch et al., 2004). DSi retention has been predominately assessed for
individual lakes and reservoirs (Cook et al., 2010, Triplett et al., 2008), and to a smaller degree
in rivers and oceans (Humborg et al., 2008; Laruelle et al., 2009; Le et al., 2010; Durr et al.,
2011). According to a model of Si mass balances executed by Laruelle et al. (2009), a
discrepancy of 96 Mt year-1
of silica was noticed between DSi entering fluvial systems and DSi
export. Anthropogenic induced retention can influence rives by decreasing Si export by as much
as 91% for specific catchments, and 18% globally (Harrison et al., 2012; Lauerwald et al., 2012).
Durr et al. (2011) assumed a natural DSi retention within lakes and wetlands to be 16 ± 8%.
Taking retention into consideration more reasonable DBSi fluxes of 6.69, 5.23, 1.79, and 2.75
Mt year-1
can be produced for the Pliocene, Miocene, Oligocene and Eocene, respectively. A
higher retention rate due to anthropogenic interference suggests a more precise modern DBSi
export of 2.11 Mt year-1
.
5.4 Global Impact
As land cover changes through time, global biogenic silica inputs have been seen to change and
in turn alter oceanic Si concentrations. Positive changes in marine and riverine silica
concentrations would have been beneficial for diatom growth, leading to diversification. The
doubling of silica flux from the Oligocene to Miocene, on account of coniferous forest and
grassland expansion, can be a potential trigger leading up to such a positive change. Although
total oceanic concentrations increase minutely with this change in Miocene vegetation, more
localized pulses in high Si concentrations, such as those near river deltas, are where we would
expect to see diatom radiation occurring.
Global fluxes estimated from the Eocene to Pliocene were completed assuming modern
environments, on which the model was constructed, are comparable. Consistency between
precipitation, discharge, lithology and soil types must be present through all time periods. The
results of this case study suggest that oceanic silica concentrations will increase as coniferous
forests and grasslands expand, which will invariably impact diatom evolution. Currently oceans
Page 68
55
are ubiquitously undersaturated in silica and changes in land cover (a result of deforestation
and/or climate effects) or pulse events could offset this, triggering another diatom radiation event
or dieback.
Page 69
56
0.385Table 12. Global area coverage (in ha) by various land cover types from the Eocene to
Pliocene
Land Cover Pliocene
Area
(ha 106)
Miocene
Area
(ha 106)
Oligocene
Area
(ha 106)
Eocene Area
(ha 106)
Rainforest 1841 2251 8486 9288
Deciduous 5047 9850 47314 80153
Grassland 6262 5011 1089 -
Conifer 15934 11904 - -
Desert 11305 8455 - -
Total 40389 37471 56889 89441
Table 13. Total global flux of dissolved biogenic silica ( ) to oceans from the Eocene to
Pliocene and resulting oceanic biogenic silica concentrations (Tsi ). Using an ocean volume
of 1.3 billion km3.
(Kg/yr)(106)
Land Cover Pliocene Miocene Oligocene Eocene
Rainforest 70 86 326 357
Deciduous 194 378 1819 3082
Grassland 565 452 98 -
Conifer 7536 5630 - -
Total 8367 6548 2244 3439
Tsi (umol/L) 0.00229 0.00179 0.00061 0.00094
Page 70
57
Figure 9. Reconstruction of Eocene (55 Mya) land cover type and distribution. After Fine
and Ree, 2006.
.
Figure 10. Reconstruction of Oligocene (27 Mya) land cover type and distribution. After
Fine and Ree, 2006 and Lunt et al., 2007.
Page 71
58
Figure 11. Reconstruction of Miocene (11 Mya) land cover type and distribution. After
Pound et al., 2011.
Figure 12. Reconstruction of Pliocene (3 Mya) land cover type and distribution. After
Haywood et al., 2004.
Page 72
59
Chapter 6
6.0 Conclusions and Future Directions of TBSi Cycles
Information concerning the terrestrial silica cycle is greatly underrepresented in literature. This is
particularly true when discussing the processes through which biogenic silica operates. Through
the construction of a terrestrial biogenic silica model, this study attempts to explicate fluxes and
storages of biogenic silica within the biosphere. The model created (Equation 2), quantifies the
amount of biogenic silica leaving terrestrial systems as a riverine flux. The riverine flux is a
function of biogenic silica leaching through soils (LF), the amount of silica dissolved within soils
(DF), net primary productivity of vegetation within a catchment (NPP), and finally the biogenic
silica content within vegetation of a catchments as percent dry weight (%BSi).
Applying this model to four land cover classes within the United States showed that vegetation
composition influences riverine silica concentrations. Catchments dominated by coniferous
forests showed to have the largest biogenic silica flux, attributed to high dissolution of
coniferous phytoliths. Grasslands showed to have the second largest BSi flux, but several times
lower than that of coniferous drainages. Low grassland BSi flux can be attributed to low
dissolution rates of grassland phytoliths, a characteristic of land cover class. Deciduous forest
riverine flux was the third lowest, with wetlands having the lowest biogenic Si export. Although
land cover characteristics influenced the amount of silica leaving these systems, the extent to
which land cover directly affects export is dependent upon physical parameters such as retention
and soil porosity.
Using the constructed terrestrial BSi model global estimates of biogenic silica export are of 0.42
Tmol Si year-1
. This value is comparable to Treguer and De La Rocha’s (2012) estimate of 1.1
Tmol Si year-1
. Global BSi flux for four time periods was recreated, demonstrating change in
dissolved terrestrial Si export with variation in vegetation cover. During the Eocene and
Oligocene DBSi export was approximately 7000 kg yr-1
, this values increased substantial during
the Miocene to 17900 kg yr-1
. This change in DBSi export is concurrent with the expansion of
grasslands and dominance of coniferous forests. To address the question of whether this rise of
terrestrial DBSi export lead to the triggering of diatom diversification more work is needed.
Page 73
60
Suggested future work would be to study of DBSi contributions of localized regions such as river
outlets to determine if multiple geographically isolated regions of diatom radiation are possible.
Although terrestrial BSi has been established as being an essential component of the carbon
cycle and driver of marine productivity, few attempts have been made to model the movement of
biogenic silica within the terrestrial sphere. Those studies that address similar processes are set
for regional environments and have minimal application for use as global terrestrial BSi cycling
models. Even with these localized examples insufficient data exists for tropical savannah,
wetlands, shrublands, and tropical dry forest systems. More work is required prying into these
less understood BSi systems and the creation of terrestrial BSi models need to be further
explored and improved upon. New geochemical and isotopic tracers have allowed for the
development of new methodologies in quantifying the influence of BSi in terrestrial
environments. These include cosmogenic and stable isotopes of Si and Ge/Si ratios, which will
hopefully be used in future studies to further expand upon our knowledge of BSi as both a
proponent of diatom diversification and terrestrial cycling. With the use of new technology and
methodologies such as Ge/Si ratios, Si isotopes and establishment of simple models, there is
upcoming promise in understanding the terrestrial Si cycle.
Page 74
61
References
Acreman, M. C., Fisher, J., Stratford, C. J., Mould, D. J., Mountford, J. O. 2007. Hydrological
science and wetland restoration: some case studies from Europe. Hydrology and Earth System
Sciences 11, 158-169.
Alexandre, A., Bouvet, M., Abbadie, L. 2011. The role of savannas in the terrestrial Si cycle: A
case-study from Lamto, Ivory Coast. Global and Planetary Change 78, 162-169.
Alexandre, A., Meunier, J. D., Colin, F., Koud, J. M. 1997. Plant impact on the biogeochemical
cycle of silicon and related weathering processes. Geochimica et Cosmochimica Acta 61, 677-
682.
Banowet, G., Casier, M., El-Nashaar, H., Griffith, S., and Vogel, K., 2009. Genotypic variability
in mineral composition of switchgrass. Bioresource Technology, 100, 1809-1814.
Bartoli, F. 1983. The biogeochemical cycle of silica in two temperate forest ecosystems. Bulletin
of the Ecological Society of America 35, 469-476.
Bernard, C. Y., Dürr, H. H., Heinze, C., Segschneider, J., Maier-Reimer, E. 2011. Contribution
of riverine nutrients to the silicon biogeochemistry of the global ocean–a model
study. Biogeosciences 8, 551-564.
Bezeau, L.M., Johnston, A., Smoliak, S. 1966. Silica and protein content of mixed prairie and
fescue grassland vegetation and its relationship to the incidence of silica urolithiasis. Canadian
Journal of Plant Science 46, 625-631.
Blank, R. R., Allen, F., Young, J. A. 1994. Growth and elemental content of several sagebrush-
steppe species in unburned and post-wildfire soil and plant effects on soil attributes. Plant and
Soil 164, 35-41.
Blecker, S. W., King, S. L., Derry, L. A., Chadwick, O. A., Ippolito, J. A., Kelly, E. F. 2007. The
ratio of germanium to silicon in plant phytoliths: quantification of biological discrimination
under controlled experimental conditions. Biogeochemistry 86, 189-199.
Page 75
62
Blecker, S.W., McCulley, R.L., Chadwick, O.A., Kelly, E.F. 2006. Biogenic cycling of silica
across and grassland bioclimosequence. Global Biogeochemical Cycles 20, 1-11.
van Cappellen, P., Qui, L. 1997. Biogenic silica dissolution in sediments of the Southern Ocean.
I Solubility. Deep Sea Research 11, 1109-1128.
Capellacci, S., Battocchi, C., Casabianca, S., Giovine, M., Bavestrello, G., Penna, A. 2012.
Bioavailability of different chemical forms of dissolved silica can affect marine diatom
growth. Marine Ecology 34, 103-111.
Carbonnel, V., Lionard, M., Muylaert, K., Chou, L. 2009. Dynamics of dissolved and biogenic
silica in the freshwater reaches of a macrotidal estuary (The Scheldt,
Belgium). Biogeochemistry 96, 49-72.
Carey, J.C., Fulweiler, R.W. 2012. The terrestrial silica pump. PLoS ONE 7, 1-4.
Carnelli, A.L., Madella, M., Theurillat, J.P. 2001. Biogenic silica production in selected alpine
plant species and plant communities. Annals of Botany 87, 425-434.
Casler, M.D., Jung, H.G. 2006. Relationships of fibre, lignin, and phenolics to in vitro fibre
digestibility in three perennial grasses. Animal Feed Science and Technology 125, 151-161.
Cocker, K. M., Evans, D. E., Hodson, M. J. 1998. The amelioration of aluminium toxicity by
silicon in higher plants: Solution chemistry or an in planta mechanism? Physiologia
Plantarum 104, 608-614.
Conley, D. J. 2002. Terrestrial ecosystems and the global biogeochemical silica cycle. Global
Biogeochemical Cycles 16, 68-1.
Cook, P. L., Aldridge, K. T., Lamontagne, S., Brookes, J. D. 2010. Retention of nitrogen,
phosphorus and silicon in a large semi-arid riverine lake system. Biogeochemistry 99, 49-63.
Cooke, J., Leishman, M. R. 2011. Is plant ecology more siliceous than we realise? Trends in
plant science 16, 61-68.
Page 76
63
Cornelis, J.T., Ranger, J., Iserentant, A., Delvaux, B. 2010. Tree species impact the terrestrial
cycle of silicon through various uptakes. Biogeochemistry 97, 231-245.
Darmody, R., David, M., Mitchell, C., Tsai, J., Voigt, T., and Woli, K., 2011. Evaluating silicon
concentrations in biofuel feedstock crops Miscanthus and switchgrass. Biomass and Bioenergy,
35, 2807-2813.
Datnoff L.E., Rodrigues F.A., Seebold K.W. 2007. Silicon and plant disease. In Mineral
Nutrition and Plant Disease, pp. 233–246. Eds L.E. Datnoff, W.H. Elmer and D.M. Huber. St.
Paul, MN, USA: The American Phytopathological Society.
David, F. 1984. The book of Bamboo. Sierra Club Books.
De Bakker, N. V. J., Hemminga, M. A., Van Soelen, J. 1999. The relationship between silicon
availability, and growth and silicon concentration of the salt marsh halophyte Spartina
anglica. Plant and Soil 215, 19-27.
Delaney, M.L. 1998. Phosphorus accumulation in marine sediments and the oceanic phosphorus
cycle. Global biogeochemical Cycles 12, 563-572.
Delvigne, C., Opfergelt, S., Cardinal, D., Delvaux, B., Andre, L. 2009. Distinct silicon and
germanium pathways in the soil‐plant system: Evidence from banana and horsetail. Journal of
Geophysical Research: Biogeosciences 114, 2005–2012.
Derry, L. A., Kurtz, A. C., Ziegler, K., Chadwick, O. A. 2005. Biological control of terrestrial
silica cycling and export fluxes to watersheds. Nature 433 , 728-731.
Ding, T.P., Zhou, J.X., Wan, D.F., Chen, Z.Y., Wang, C.Y., Zhang, F. 2009. Silicon isotope
fractionation in bamboo and its significance to the biogeochemical cycle of silicon. Geochimica
et Cosmochimica Acta 72, 1381-1395.
Dixit, S., Van Cappellen, P., van Bennekom, A. J. 2001. Processes controlling solubility of
biogenic silica and pore water build-up of silicic acid in marine sediments. Marine Chemistry 73,
333-352.
Page 77
64
Dodson, E.K., Root, H.T. 2013. Conifer regeneration following stand-replacing wildfire varies
along an elevation gradient in a ponderosa pine forest, Oregon, USA. Forest Ecology and
Management 302, 163-170.
Dowsett, H. J., Chandler, M. A., Cronin, T. M., Dwyer, G. S. 2005. Middle Pliocene sea surface
temperature variability. Paleoceanography 20, 1-8.
Dove, P. M., & Elston, S. F. 1992. Dissolution kinetics of quartz in sodium chloride solutions:
Analysis of existing data and a rate model for 25 C. Geochimica et Cosmochimica Acta 56,
4147-4156.
Dove, P., Han, N., Wallace, A.F., De Yoreo, J. 2007. Kinetics of amorphous silica dissolution
and the paradox of the silica polymorphs. Proceedings of the National Academy of Sciences 105,
9903-9908.
Drees, L. R., Wilding L. P., Smeck N. E., Senkayi A. L. 1989. Silica in soils: Quartz and
disordered silica polymorphs, in Minerals in Soils Environments, edited by J. B. Dixon and S. B.
Weed, pp. 913– 974, Soil Science Society of America, Madison Wis.
Drever, J.I. 1994. The effect of land plants on weathering rates of silica minerals. Geochimica et
Cosmochimica Acta 58, 2325-2332.
Dürr, H. H., Meybeck, M., Hartmann, J., Laruelle, G. G., Roubeix, V. (2011). Global spatial
distribution of natural riverine silica inputs to the coastal zone. Biogeosciences 8, 597-620.
Egge, J.K., Aksnes, D.L. 1992. Silicate as regulating nutrient in phytoplankton competition.
Marine Ecology Progress Series 83, 281-289.
Epstein, E. 1994. The anomaly of silicon in plant biology. Proceedings of the National Academy
of Sciences 91, 11-17.
Falkowski, P.G., Katz, M.E., Knoll, A.H., Quigg, A., Raven, J.A., Schofield, O., Taylor, F.J.R.
2004. The evolution of modern eukaryotic phytoplankton. Science 305, 354-360.
Fauteux, F., Remus-Borel, W., Menzies, J.G., Belanger, R.R. 2005. Silicon and plant disease
resistance against pathogenic fungi. FEMS Microbiology Letters 249, 1-6.
Page 78
65
Fine, P. V., Ree, R. H. 2006. Evidence for a time‐integrated species‐area effect on the latitudinal
gradient in tree diversity. The American Naturalist 168, 796-804.
Fishkis, O., Ingwersen, J., Lamers, M., Denysenko, D., Streck, T. 2010. Phytolith transport in
soil: a field study using fluorescent labelling. Geoderma 157, 27-36.
Fraysse, F., Pokrovsky, O.S., Schott, J., Meunier, J.D. 2009. Surface chemistry and reactivity of
plant phytoliths in aqueous solutions. Chemical Geology 258, 197-206.
Fraysse, F., Pokrovsky, O.S., Schott, J., Meunier, J.D. 2006. Surface properties, solubility and
dissolution kinetics of bamboo phytoliths. Geochimica et Cosmochimica Acta 70, 1939-1951.
Friedl, G., Teodoru, C., Wehrli, B. 2004. Is the Iron Gate I reservoir on the Danube River a sink
for dissolved silica?. Biogeochemistry 68, 21-32.
Gali-Muhtasib, H.U., Smith, C.C., Higgins, J.J. 1992. The effect of silica in grasses on the
feeding behavior of prairie Vole, Microtus Ochrogaster. Ecological Society of America 73,
1724-1729.
Garbuzov, M., Reidinger, S., Hartley, S.E. 2011. Interactive effects of plant-available soil silicon
and herbivory on competition between two grass species. Annals of Botany 108, 1355-1363.
Garrison, T.S. 2005. Oceanography: An invitation to marine science. Thompson Brooks/Cole,
pp4.
Garvin, C.J. 2006. An exploratory study of the terrestrial biogeochemical silicon cycle at a
forested watershed in Northern Vermont. PhD diss., Cornell University.
Geis, J.W. 1978. Biogenic silica in selected species of deciduous angiosperms. Soil Science 116,
113-119.
Geis, J. W. 1978. Biogenic opal in three species of Gramineae. Annals of Botany 42, 1119-1129.
Gerard, F., Francois, M., Ranger, J. 2002. Processes controlling silica concentration in leaching
and capillary soils solutions of an acidic brown forest soil (Rhone, France). Geoderma 107, 197-
226.
Page 79
66
Gholz, H.L. 1982. Environmental limits on aboveground net primary production, leaf area, and
biomass in vegetation zones of the Pacific Northwest. Ecology 63, 469-481.
Gieslet, R., Ilvesniemi, H., Nyberg, L., van Hees, P., Starr, M., Bishop, K., Kareinen, T.,
Lundstrom, U.S. 2000. Distribution and mobilization of Al, Fe, and Si in three podzolic soil
profiles in relation to the humus later. Geoderma 94, 249-263.
Given, P.H., Miller, R.N. 1985. Distribution of forms f sulfur in peats from saline environments
in the Florida everglades. International Journal of Coal Geology 5, 397-409.
Greenwood, J.E., Trusdale, V.W., Rendell, A.R. 2011. Biogenic silica dissolution in seawater- in
vitro chemical kinetics. Progress in Oceanography 48, 1-23.
Gregg, W.W., Casey, N.W. 2007. Modeling coccolithophores in the global oceans. Deep Sea
Research 54, 447-477.
Guntzer, F., Keller, C., and Meunier, J., 2012. Benefits of plant silicon for crops: a review.
Agronomy for sustainable development, 32, 1, 201-213.
Harrison, J. A., Frings, P. J., Beusen, A. H., Conley, D. J., McCrackin, M. L. 2012. Global
importance, patterns, and controls of dissolved silica retention in lakes and reservoirs. Global
Biogeochemical Cycles 26, 1-12.
Haywood, A. M., Valdes, P. J. 2004. Modelling Pliocene warmth: contribution of atmosphere,
oceans and cryosphere. Earth and Planetary Science Letters 218, 363-377.
Hicke, J.A., Asner, G.P., Randerson, J.T., Tucker, C., Los, S., Birdsey, R., Jenkins, J.C., Field,
C. 2002. Trends in North American net primary productivity derived from satellite observations,
1982–1998. Global Biogeochemical Cycles 16, 1029-1039.
Hill, R., Pickering, C. 2009. Differences in resistance of three subtropical vegetation types to
experimental trampling. Journal of Environmental Management 90, 1305-1312Hodson, M. J., &
Evans, D. E. 1995. Aluminium/silicon interactions in higher plants. Journal of Experimental
Botany 46, 161-171.
Page 80
67
Hinsinger, P., Barros, O., Benedetti, M.F., Noack, Y., Callot, G. 2001. Plant-induced weathering
of a basaltic rock: Experimental evidence. Geochimica et Cosmochimica Acta 65, 137-152.
Hodson, M.J., Sangster, A.G. 1999. Aluminium/silicon interaction in conifers. Journal of
Inorganic Biochemistry 76, 89-98.
Hodson, M.J., White, P.J., Mead, A., Broadley, M.R. 2005. Phylogenetic variation in the silicon
composition of plants. Annals of Botany 96, 1027-1046.
Horst, W.J., Fecht, M., Naumann, A., Wissemeier, A.H., Maier, P. Physiology of manganese
toxicity and tolerance in Vigna unguiculata (L.) Walp. Journal of Plant Nutrition and Soil
Science 162, 293-274.
Horst, W.J., Marschner, H. 1978. Effect of silicon on manganese tolerance of bean plants
(Phaseolus vulgaris L.). Plant and Soil 50, 1-3.
Humborg, C., Smedberg, E., Medina, M. R., Mörth, C. M. 2008. Changes in dissolved silicate
loads to the Baltic Sea—the effects of lakes and reservoirs.Journal of Marine systems 73, 223-
235.
Johnston, A., Bezeau, L. M., Smoliak, S. 1967. Variation in silica content of range
grasses. Canadian Journal of Plant Science 47, 65-71.
Jorgensen, E.G. 1953. Silicate assimilation by diatoms. Physiologia Plantarum 6, 301-315.
Kalisz, P. J., Stone, E. L. 1984. The longleaf pine islands of the Ocala National Forest, Florida: a
soil study. Ecology, 1743-1754.
Kamykowski, D., Zentara, S.-J., Morrison, J.M., Switzer, A.C., 2002. Dynamic global patterns of
nitrate, phosphate, silicate, and iron availability and phytoplankton community composition from
remote sensing data. Global Biogeochemical Cycles 16, 1-29.
Kanto, T., Miyoshi, A., Ogawa, T., Maekawa, K., Aino, Masataka. 2006. Supressive effect of
liquid potassium silicate on powdery mildew of strawberry in soil. Journal of General Plant
Pathology 72, 137-142.
Page 81
68
Keller, C., Meunier, J., Miche, H., and Rizwan, M., 2012. Effectof silicon on reducing cadmium
toxicity in durum wheat (Triticum turgidum L. cv. Claudio W.) grown in a soil with aged
contamination. Journal of Hazardous Materials, 209-210, 326-334.
Kelly, E. F., Chadwick, O.A., Hilinski, T.E. 1998. The effect of plants on mineral
weathering, Biogeochemistry 42, 21–53.
Klein, R.L., Geis, J.W. 1978. Biogenic silica in the Pinaceae. Soil Science 126, 145-156.
Koch, R. W., Guelda, D. L., Bukaveckas, P. A. 2004. Phytoplankton growth in the Ohio,
Cumberland and Tennessee Rivers, USA: inter-site differences in light and nutrient
limitation. Aquatic Ecology 38, 17-26.
Konhauser, K. O., Urrutia, M. M. 1999. Bacterial clay authigenesis: a common biogeochemical
process. Chemical Geology, 16, 399-413.
Kurtz, A. C., Derry, L. A., Chadwick, O. A. 2002. Germanium-silicon fractionation in the
weathering environment. Geochimica et Cosmochimica Acta 66, 1525-1537.
Kurz, I., O’Reilly, C. D., Tunney, H. 2006. Impact of cattle on soil physical properties and
nutrient concentrations in overland flow from pasture in Ireland.Agriculture, ecosystems &
environment 113, 378-390.
Laclau, J. P., Sama-Poumba, W., Nzila, J. D. D., Bouillet, J. P., Ranger, J. 2002. Biomass and
nutrient dynamics in a littoral savanna subjected to annual fires in Congo. Acta Oecologica 23,
41-50.
Lanning, F. C., Eleuterius, L. N. 1987. Silica and ash in native plants of the central and
southeastern regions of the United States. Annals of Botany 60, 361-375.
Lanning, F. C., Eleuterius, L. N. 1989. Silica deposition in some C3 and C4 species of grasses,
sedges and composites in the USA. Annals of Botany 64, 395-410.
Laruelle, G. G., Roubeix, V., Sferratore, A., Brodherr, B., Ciuffa, D., Conley, D. J.,Van
Cappellen, P. 2009. Anthropogenic perturbations of the silicon cycle at the global scale: Key role
of the land‐ocean transition. Global biogeochemical cycles 23.
Page 82
69
Lasaga, A. 1984. Chemical kinetics of water-rock interactions. Journal of Geophysical Research.
89, 4009-4025.
Lauerwald, R., Hartmann, J., Moosdorf, N., Dürr, H. H., Kempe, S. 2012. Retention of dissolved
silica within the fluvial system of the conterminous USA. Biogeochemistry, 1-23.
Lewin, J.C. 1961. The dissolution of silica from diatom walls. Geochimica et Cosmochimica
Acta 21, 182-198.
Le, T. P. Q., Gilles, B., Garnier, J., Sylvain, T., Denis, R., Anh, N. X., Minh, C. V. 2010.
Nutrient (N, P, Si) transfers in the subtropical Red River system (China and Vietnam): modelling
and budget of nutrient sources and sinks.Journal of Asian Earth Sciences 37, 259-274.
Li, Z., Song, Z., Li, B. 2013a. The production and accumulation of phytolith-occluded carbon in
Baiyangdian reed wetland of China. Applied Geochemistry 37, 117-124.(b)
Li, Z., Song, Z., Parr, J. F., Wang, H. 2013b. Occluded C in rice phytoliths: implications to
biogeochemical carbon sequestration. Plant and Soil 1, 1-9.(a)
Litton, C., M, Raich, J., W, Ryan, M., G. 2007.Carbon allocation in forest ecosystems. Global
Change Biology 13: 2089–2109
.Lorenzen, B., Brix, H., Mendelsson, I.A., McKee, K.L., Miao, S.L. 2001. Growth, biomass
allocation and nutrient use efficiency in Cladium jamaicense and Typha domingensis as affected
by phosphorus and oxygen availability. Aquatic Botany 70, 117–133.
Loucaides, S., Behrends, T., Van Cappellen, P. 2010. Reactivity of biogenic silica: Surface
versus bulk charge density. Geochimica et Cosmochimica Acta 74, 517-530.
Loucaides, S., Van Cappellen, P., Behrends, T. 2008. Dissolution of biogenic silica from land to
ocean: Role of salinity and pH. Association for the Sciences of Limnology and Oceanography
53, 1614-1621.
Lucas, Y., Luizao, F.J., Chauvel, A., Rouiller, J., Nahon, D. 1993. The relation between
biological activity of the rain forest and mineral composition of soils, Science, 260, 521 – 523.
Page 83
70
Lunt, D. J., Ross, I., Hopley, P. J., Valdes, P. J. 2007. Modelling late Oligocene C grasses and
climate. Palaeogeography, Palaeoclimatology, Palaeoecology 251, 239-253.
Luu, T. N. M., Garnier, J., Billen, G., Le, T. P. Q., Nemery, J., Orange, D., Le, L. A. 2012. N, P,
Si budgets for the Red River Delta (northern Vietnam): how the delta affects river nutrient
delivery to the sea. Biogeochemistry 107, 241-259.
Ma, J.F., Miyake, Y. 2001. Silicon as a beneficial element for crop plants. Silicon in Agriculture
8, 17-39.
Ma JF, Takahashi E. 2002. Soil, fertilizer, and silicon research in Japan. Elsevier: Amsterdam.
Ma, J.F., Yamaji, N. 2006. Silicon uptake and accumulation in higher plants. TRENDS in Plant
Science 11, 392-397.
McCarthy, T. S., J. R. McIver, B. Cairncross, W. N. Ellery, and Karen Ellery. 1989. The
inorganic chemistry of peat from the Maunachira channel-swamp system, Okavango Delta,
Botswana. Geochimica et Cosmochimica Acta 53, 1077-1089.
Martin‐Jézéquel, V., Hildebrand, M., Brzezinski, M. A. 2000. Silicon metabolism in diatoms:
implications for growth. Journal of Phycology 36, 821-840.
Manabde, M.., Sivapalan, M. 2001. Linking space time variability of river runoff and rainfall
fields: a dynamic approach. Advances in Water Resources 24, 1001-1014.
Markewitz, D., Richter, D. 1998. The bio in aluminum and silica geochemistry. Biogeochemistry
42, 235-252.
Massey, F.P., Ennos, A.R., Hartley, S.E. 2007. Herbivore specific induction of silica based plant
defences. Oecologia 152, 667-683.
Melzer, S. E., Chadwick, O. A., Hartshorn, A. S., Khomo, L. M., Knapp, A. K., Kelly, E. F.
2012. Lithologic controls on biogenic silica cycling in South African savanna
ecosystems. Biogeochemistry 108, 317-334.
Meunier, J.D., Colin, F., Alarcon, C. 1999. Biogenic silica storage in soils. Geology 27, 835-838.
Page 84
71
Milesi, C., Elvidge, C.D., Nemani, R.R., Running, S.W. 2003. Assessing the impact of urban
land development on net primary productivity in the southeastern United States. Remote Sensing
of Environment 86, 401-410.
Moulton, K.L., Berner, R.A. 1998. Quantification of the effect of plants on weathering: Studies
in Iceland. Geology 26, 895-898.
Nahlik, A. M., Mitsch, W. J. 2008. The effect of river pulsing on sedimentation and nutrients in
created riparian wetlands. Journal of environmental quality 37, 1634-1643.
Nelson, D.M., DeMaster, D.J., Dunbar, R.B., Smith, W.O. 1996. Cycling of organic carbon and
biogenic silica in the Southern Ocean: Estimates of water-column and sedimentary fluxes on the
Ross Sea continental shelf. Journal of Geophysical Research 101, 18519-18532.
Neubauer, E., Schenkeveld, W. D., Plathe, K. L., Rentenberger, C., von der Kammer, F.,
Kraemer, S. M., Hofmann, T. 2013. The influence of pH on iron speciation in podzol extracts:
iron complexes with natural organic matter, and iron mineral nanoparticles. Science of The Total
Environment 461, 108-116.
Nikodem, A., Kodešová, R., Drábek, O., Bubeníčková, L., Borůvka, L., Pavlů, L., Tejnecký, V.
2010. A numerical study of the impact of precipitation redistribution in a beech forest canopy on
water and aluminum transport in a Podzol. Vadose Zone Journal 9, 238-251.
Norby, R.J., Hanson, P.J., O’Neill, E.G., Tschaplinski, T.J., Weltzin, J.F., Hansen, R.A., Weixin,
C., Wullschleger, S.D., Gunderson, C.A., Edwards, N.T., Johnson, D.W. 2002. Net primary
productivity of a CO2 enriched deciduous forest and the implications for carbon storage.
Ecological Applications 12, 1261-1266.
Norris, A. R., Hackney, C. T. 1999. Silica content of a mesohaline tidal marsh in North
Carolina. Estuarine, Coastal and Shelf Science 49, 597-605.
O’Green, A.,T. 2012. Soil water dynamics. Nature Education Knowledge 3, 1-12.
Okuda, A., Takahashi.1962. Effect of silicon supply on the injuries due to excessive amount of
Fe, Mn, Cu, As, Al, Co of barley and rice plant. Soil Science and Plant Nutrition 33, 1-8.
Page 85
72
Opalinska, B., Cowling, S. A. 2013. Using carbon economics of tree height to estimate
evolutionary timing of cold tolerance in conifers. Palaeogeography, Palaeoclimatology,
Palaeoecology 392, 495-501.
Opdekamp, W., J. Teuchies, D. Vrebos, J. Chormanski, J. Schoelynck, and R. Van Diggelen. P.,
M. 2012. Tussocks: biogenic silica hot-spot in a riparian wetland. Wetlands 32, 1115-1124.
Ovington, J. D. 1956. The composition of tree leaves. Forestry 29, 22-28.
Opfergelt, S., Cardinal, D., André, L., Delvigne, C., Bremond, L., Delvaux, B. 2010. Variations
of δ30 Si and Ge/Si with weathering and biogenic input in tropical basaltic ash soils under
monoculture. Geochimica et Cosmochimica Acta, 74, 225-240.
Piperno, D. R., Ranere, A. J., Holst, I., Iriarte, J., Dickau, R. 2009. Starch grain and phytolith
evidence for early ninth millennium BP maize from the Central Balsas River Valley,
Mexico. Proceedings of the National Academy of Sciences 106, 5019-5024.
Pisaric, M. F. 2002. Long-distance transport of terrestrial plant material by convection resulting
from forest fires. Journal of Paleolimnology 28, 349-354.
Pokrovsky, A. L., Kamaev, V., Li, C. Y., Vardeny, Z. V., Efros, A. L., Kurdyukov, D. A.,
Golubev, V. G. 2005. Theoretical and experimental studies of metal-infiltrated opals. Physical
Review B 71, 1-5.
Pound, M. J., Haywood, A. M., Salzmann, U., Riding, J. B., Lunt, D. J.,Hunter, S. J. 2011. A
Tortonian (late Miocene, 11.61–7.25 Ma) global vegetation reconstruction. Pala ogeography,
Palaeoclimatology, Palaeoecology 300, 29-45.
Rabosky, D.L., Sorhannus, U. 2009. Diveristy dynamics of marine planktonic diatoms across the
Cenozoic. Nature 457, 183-187.
Raven, J.A. 1983. The transport and function of silicon in plants. Biological Reviews 58, 179-
207.
Rickert, D., Schluter, M., Wallmann, K. 2002. Dissolution kinetics of biogenic silica from the
water column to the sediments. Geochimica et Cosmochimica Acta 66, 439-455.
Page 86
73
Robbins, C. T., Mole, S., Hagerman, A. E., Hanley, T. A. 1987. Role of tannins in defending
plants against ruminants: reduction in dry matter digestion? Ecology 68, 1606-1615.
Rocha, A.D., Goulden, M.L. 2009. Why is marsh productivity so high? New insights from eddy
covariance and biomass measurements in a Typha marsh. Agricultural and Forest Meteorology
149, 159-168.
Ryves, D.B., Battarbee, R.W., Juggins, S., Fritz, S.C., Anderson, N.J. 2006. Physical and
chemical predictors of diatom dissolution in freshwater and saline lake sediments in North
American and West Greenland. Association for the Sciences of Limnology and Oceanography
51, 1355-1368.
Ryves, D. B., Juggins, S., Fritz, S. C., Battarbee, R. W. 2001. Experimental diatom dissolution
and the quantification of microfossil preservation in sediments. Palaeogeography,
Palaeoclimatology, Palaeoecology 172, 99-113.
Saccone, L., Conley, D.J., Koning, E., Sauer, D., Sommer, M., Kaczorek, D., Blecker, S.W.,
Kelly, E. 2007. Assessing the extraction and quantification of amorphous silica in soils of forest
and grassland ecosystems. European Journal of Soil Science 58, 1446-1459.
Saijonkari-Pahkala, K. 2008. Non-wood plants as raw material for pulp and paper. Agricultural
and Food Science 10, 101-110.
Sarmiento, J.L., Gruber, N. 2006. Ocean biogeochemical dynamics. Princeton, NJ, Princeton
University Press. pp. 526.
Schoelynck, J., Bal, K., Backx, H., Okruszko, T., Meire, P., Struyf, E. 2010. Silica uptake in
aquatic and wetland macrophytes: a strategic choice between silica, lignin and cellulose? New
Phytologist 186, 385-391.
Scurlock, J.M., Olson, R.J. 2002. Terrestrial net primary productivity – A brief history and a new
worldwide database. Environmental Reviews 10, 91-109.
Shimoyama. 1958. Effect of silicon on lodging and wind damage in rice. Report for the Research
Funds Granted by Ministry of Agriculture, Japan. pp. 82.
Page 87
74
Siever R. 1991. The silica cycle in the Precambrian. Geochimica et Cosmochimica Acta 56,
3265-3272.
Smith, G. S., Nelson, A. B., Boggino, E. J. 1971. Digestibility of forages in vitro as affected by
content of ―silica‖. Journal of Animal Science 33, 466-471.
Sommer, M., Halm, D., Weller, U., Zarei, M., Stahr, K. 2000. Lateral podzolization in a granite
landscape.Soil Science Society of America Journal 64, 1434–1442.
Sommer M., Kaczorek, D., Kuzyakov, Y., Breuer, J. 2006. Silicon pools and fluxes in soils and
landscapes-a review. Journal of Plant Nutrition and Soil Science 169, 310-329.
Song, Z., Zhao, S., Zhang, Y., Hu, G., Cao, Z., Wong, M. 2011. Plant impact on CO2
consumption by silicate weathering: The role of bamboo. Botanical Review 77, 208-213.
Strömberg, C. A. 2011. Evolution of grasses and grassland ecosystems. Annual Review of Earth
and Planetary Sciences 39, 517-544.
Struyf, E., Conley, D.J. 2009. Silica: An essential nutrient in wetland biogeochemistry. Ecology
and the Environment 7, 88-94.
Struyf, E., Van Damme, S., Gribsholt, B., Middelburg, J. J., Meire, P. 2005. Biogenic silica in
freshwater marsh sediments and vegetation (Schelde estuary, Belgium).Marine Ecology 303, 51-
60.
Struyf, E., Van Damme, S., Gribsholt, B., Bal, K., Beauchard, O., Middelburg, J. J., Meire, P.
2007. Phragmites australis and silica cycling in tidal wetlands. Aquatic botany 87, 134-140.
Treguer, P.J., De La Rocha, C.L. 2012. The world ocean silica cycle. Annual Review of Marine
Science 5, 477-501.
Treguer, P.J., Nelson, D.M., Van Bennekom, A.J., DeMaster, D.J., Leynaert, A., Queguiner, B.
1995. The silica balance of the world ocean: A reestimate. Science 268, 375- 379.
Page 88
75
Triplett, L. D., Engstrom, D. R., Conley, D. J., Schellhaass, S. M. 2008. Silica fluxes and
trapping in two contrasting natural impoundments of the upper Mississippi
River. Biogeochemistry 87(3), 217-230.
Van Cappellen, P., Dixit, S., van Beusekom, J. 2002. Biogenic silica dissolution in the oceans:
Reconciling experimental and field‐based dissolution rates. Global Biogeochemical Cycles 16,
23-1.
Van Iersel, M.W., 2003. Carbon use efficiency depends on growth respiration, maintenance
respiration, and relative growth rate. A case study with lettuce. Plant, Cell, & Environment 26,
1441–1449.
van der Valk, A. G., Warner, B. G. 2009. The development of patterned mosaic landscapes: an
overview. Plant Ecology 200, 1-7.
Viaroli, P., Nizzoli, D., Pinardi, M., Rossetti, G., Bartoli, M. 2013. Factors Affecting Dissolved
Silica Concentrations, and DSi and DIN Stoichiometry in a Human Impacted Watershed (Po
River, Italy). Silicon 5, 101-114.
White, Art F., Davison V. Vivit, Marjorie S. Schulz, Tom D. Bullen, Evett, R., Jugdeep, A.
2012. Biogenic and pedogenic controls on Si distributions and cycling in grasslands of the Santa
Cruz soil chronosequence, California. Geochimica et Cosmochimica Acta 94, 72-94.
Wilding, L.P., Drees, L.R., 1974. Contributions of forest opal and associated crystalline phases
to fine silt and clay fractions of soils. Clays and Clay Minerals. 22, 295 – 306.
Yeo, A.R., Flowers, S.A., Rao, G., Welfare, K., Senanyake, N., Flowers, T.J. 1999. Silicon
reduces sodium uptake in rice (Oryza sativa L.) in saline conditions and this is accounted for by
reduction in the transpirational bypass flow. Plant, Cell & Environment 22, 559-565.
Zeyen, R.J. 2002. Silicon in plant cell defenses against cereal powdery mildew disease. In
Silicon in Agriculture Conference 2, Abstracts, Tsuruoka, Japan, pp. 15-21.
Ziegler, A., Eshel, G., Rees, P. M., Rothfus, T., Rowley, D., Sunderlin, D. (2003). Tracing the
tropics across land and sea: Permian to present. Lethaia 36, 227-254.
Page 89
76
Appendix I.
Biogenic silica content of vegetation (%DW) belonging to grasslands,
wetlands, coniferous forests and deciduous forests.
Table of biogenic silica content in American grassland species as % dry weight.
Family Genus species %Bsi Reference
Poaceae Panicum virgatum 1.76 Darmody, et al., 2011
Poaceae Panicum virgatum 0.6955 Banowet, et al., 2009
Poaceae Panicum virgatum 0.8636 Banowet, et al., 2009
Poaceae Panicum virgatum 0.764 Banowet, et al., 2009
Poaceae Panicum virgatum 0.8724 Banowet, et al., 2009
Poaceae Panicum virgatum 0.6958 Banowet, et al., 2009
Poaceae Panicum virgatum 0.7197 Banowet, et al., 2009
Poaceae Panicum virgatum 0.1383 Banowet, et al., 2009
Poaceae Panicum virgatum 1.1219 Banowet, et al., 2009
Poaceae Panicum virgatum 1.3956 Banowet, et al., 2009
Poaceae Panicum virgatum 1.2526 Banowet, et al., 2009
Poaceae Panicum virgatum 1.2543 Banowet, et al., 2009
Poaceae Panicum virgatum 1.3033 Banowet, et al., 2009
Poaceae Panicum virgatum 0.886 Banowet, et al., 2009
Poaceae Panicum virgatum 1.0296 Banowet, et al., 2009
Poaceae Panicum virgatum 1.0501 Banowet, et al., 2009
Poaceae Panicum virgatum 1.131 Banowet, et al., 2009
Poaceae Panicum virgatum 0.9574 Banowet, et al., 2009
Poaceae Panicum virgatum 0.8369 Banowet, et al., 2009
Poaceae Panicum virgatum 0.7497 Banowet, et al., 2009
Poaceae Panicum virgatum 0.4908 Banowet, et al., 2009
Poaceae Panicum virgatum 0.5516 Banowet, et al., 2009
Poaceae Panicum virgatum 0.4535 Banowet, et al., 2009
Poaceae Panicum virgatum 0.3818 Banowet, et al., 2009
Poaceae Panicum virgatum 0.6529 Banowet, et al., 2009
Poaceae Panicum virgatum 0.4533 Banowet, et al., 2009
Poaceae Panicum virgatum 0.9736 Banowet, et al., 2009
Poaceae Panicum virgatum 1.0909 Banowet, et al., 2009
Poaceae Panicum virgatum 1.2969 Banowet, et al., 2009
Poaceae Panicum virgatum 1.2441 Banowet, et al., 2009
Poaceae Panicum virgatum 0.8945 Banowet, et al., 2009
Poaceae Panicum virgatum 0.8146 Banowet, et al., 2009
Poaceae Triticum aestivum 2.455 Guntzer, et al., 2012
Page 90
77
Poaceae Zea mays 0.827 Guntzer, et al., 2012
Poaceae Triticum aestivum 2.455 Guntzer, et al., 2012
Poaceae Hordeum vulgare 1.824 Guntzer, et al., 2012
Poaceae Triticum turgidum 0.719 Keller, et al., 2012
Poaceae Elymus elymoides 1.6 Blank et al., 1994
Poaceae Avena sativa 2.08 Ma and Takahashi, 2002.
Poaceae Triticum aestivum 1.44 Ma and Takahashi, 2002
Poaceae Triticum boeoticum 2.61 Ma and Takahashi, 2002
Poaceae Triticum dicoccoides 1.33 Ma and Takahashi, 2002
Poaceae Triticum percicum 1.7 Ma and Takahashi, 2002
Poaceae - - 5 Pahkala, 2001
Poaceae - - 4.5 Pahkala, 2001
Poaceae - - 4.5 Pahkala, 2001
Poaceae - - 2 Pahkala, 2001
Poaceae Secale cereale 1.4 Robbins et al., 1987
Poaceae Triticum aestivum 2.7 Robbins et al., 1987
Poaceae Achnatherum hymenoides 2 Blank et al. 1994
Poaceae Bromus tectorum 1.5 Blank et al. 1994
Poaceae Agropyron smithii 2.52 Bezeau et al. 1966
Poaceae Agropyron subsecundum 2.23 Bezeau et al. 1966
Poaceae Bromus pumpellianus 1.86 Bezeau et al. 1966
Poaceae Calamagrostis inexpensa 3.8 Bezeau et al. 1966
Poaceae Calamagrostis rubescens 3.29 Bezeau et al. 1966
Poaceae Danthonia intermedia 3.08 Bezeau et al. 1966
Poaceae Danthonia parryi 2.61 Bezeau et al. 1966
Poaceae Deschampsia caespitosa 1.85 Bezeau et al. 1966
Poaceae Elymus cinereus 2.1 Bezeau et al. 1966
Poaceae Elymus innovatus 2.13 Bezeau et al. 1966
Poaceae Festuca idahoensis 3.59 Bezeau et al. 1966
Poaceae Festuca scabrella 3.15 Bezeau et al. 1966
Poaceae Stipa richardsonii 2.64 Bezeau et al. 1966
Poaceae Stipa spartea 2.86 Bezeau et al. 1966
Poaceae Stipa viridula 3.4 Bezeau et al. 1966
Poaceae Bromus inermis 2.47 Bezeau et al. 1966
Poaceae Elymus junceus 2.37 Bezeau et al. 1966
Poaceae Festuca rubra 2.82 Bezeau et al. 1966
Poaceae Phleum pratense 1.59 Bezeau et al. 1966
Poaceae Andropogon gerardi 4.1 Geis, 1978
Poaceae Panicum virgatum 4.39 Geis, 1978
Poaceae Sorgastrum nutans 5.65 Geis, 1978
Poaceae Andropogon gerardi 1.98
Lanning and Eleuterius,
1987
Page 91
78
Poaceae Andropogon scoparius 6.24
Lanning and Eleuterius,
1987
Poaceae Panicum virgatum 4.88
Lanning and Eleuterius,
1987
Poaceae Sorgastrum nutans 4.683
Lanning and Eleuterius,
1987
Poaceae Stipa spartea 7.15
Lanning and Eleuterius,
1987
Poaceae - - 2.5 Pahkala, 2001
Poaceae - - 3.5 Pahkala, 2001
Poaceae Sporobolus cryptandrus 3.5 Smith et al., 1971
Poaceae Lycurus phleoides 4.16 Smith et al., 1971
Poaceae Boutelous gracilis 5.72 Smith et al., 1971
Poaceae Panicum obtusum 4.41 Smith et al., 1971
Poaceae Boutelous curtipendula 5.64 Smith et al., 1971
Poaceae Hilaria jamesii 5.18 Smith et al., 1971
Poaceae Boutelous hirsuta 6.68 Smith et al., 1971
Poaceae Muhlenbergia richardsonis 7.34 Smith et al., 1971
Poaceae Festuca scabrella 2.79 Johnston et al., 1966
Poaceae Festuca scabrella 2.19 Johnston et al., 1966
Poaceae Festuca scabrella 3.41 Johnston et al., 1966
Poaceae Festuca scabrella 3.61 Johnston et al., 1966
Poaceae Festuca scabrella 3.9 Johnston et al., 1966
Poaceae Festuca scabrella 2.4 Johnston et al., 1966
Poaceae Festuca scabrella 2.94 Johnston et al., 1966
Poaceae Festuca scabrella 4.72 Johnston et al., 1966
Poaceae Festuca scabrella 4.33 Johnston et al., 1966
Poaceae Festuca scabrella 3.35 Johnston et al., 1966
Poaceae Festuca scabrella 3.39 Johnston et al., 1966
Poaceae Festuca scabrella 2.39 Johnston et al., 1966
Poaceae Festuca scabrella 3.73 Johnston et al., 1966
Poaceae Deschampsia caespitosa 1.86 Johnston et al., 1966
Poaceae Stipa comata 0.89 Johnston et al., 1966
Poaceae Stipa comata 1.5 Johnston et al., 1966
Poaceae Stipa comata 1.55 Johnston et al., 1966
Poaceae Stipa comata 1.08 Johnston et al., 1966
Poaceae Stipa comata 2.62 Johnston et al., 1966
Poaceae Stipa comata 1.83 Johnston et al., 1966
Poaceae Stipa comata 2.54 Johnston et al., 1966
Poaceae Stipa comata 1.13 Johnston et al., 1966
Poaceae Stipa comata 0.88 Johnston et al., 1966
Poaceae Stipa comata 1.17 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.45 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.6 Johnston et al., 1966
Page 92
79
Poaceae Bouteloua gracilis 3.24 Johnston et al., 1966
Poaceae Bouteloua gracilis 1.3 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.26 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.51 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.51 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.12 Johnston et al., 1966
Poaceae Bouteloua gracilis 1.77 Johnston et al., 1966
Poaceae Bouteloua gracilis 2.73 Johnston et al., 1966
Poaceae Bromus inermis 1.3 Robbins et al., 1987
Poaceae Bromus inermis 0.8 Robbins et al., 1987
Poaceae Bromus tectorum 2.4 Robbins et al., 1987
Poaceae Cortaderia selloana 0.04
Lanning and Eleuterius
1989
Poaceae Imperata cylindrica 1.34
Lanning and Eleuterius,
1989
Poaceae Phalaris arundinaceae 1.67 Pahkala and Pihala, 2000
Poaceae Phalaris arundinaceae 4.04 Pahkala and Pihala, 2000
Poaceae Festuca arundinaceae 0.67 Pahkala and Pihala, 2000
Poaceae Festuca arundinaceae 1.94 Pahkala and Pihala, 2000
Poaceae Festuca pratensis 0.53 Pahkala and Pihala, 2000
Poaceae Festuca pratensis 2.54 Pahkala and Pihala, 2000
- - - 0.8 Alexandre et al., 2010
- - - 1.13 Carey and Fulweiler, 2012
- - - 1.53 Carey and Fulweiler, 2012
- - - 1.37 Carey and Fulweiler, 2012
AVERAGE
2.301
Page 93
80
Table of biogenic silica content in American wetland species as % dry weight.
Family Genus Species %Bsi Reference
Poaceae Elymus virginicus 1.56 Lanning and Eleuterius, 1985
Poaceae Phragmites communis 0.6 Lanning and Eleuterius, 1985
Poaceae Ctenium aromaticum 1.41 Lanning and Eleuterius, 1985
Poaceae Ctenium aromaticum 1 Lanning and Eleuterius, 1985
- Elytrigia atherica 1.6 de Bakker, et al., 1999
Poaceae Spartina anglica 1.22 de Bakker, et al., 1999
Poaceae Festuca rubra 0.76 de Bakker, et al., 1999
Asteraceae Seriphidium maritimum 0.31 de Bakker, et al., 1999
Poaceae Puccinellia maritima 0.25 de Bakker, et al., 1999
Amaranthaceae Suaeda maritima 0.12 de Bakker, et al., 1999
Plantaginaceae Plantago maritima 0.12 de Bakker, et al., 1999
Asteraceae Aster tripolium 0.11 de Bakker, et al., 1999
Ceratophyllaceae Certophyllum demersum 2.8 Schoelynck et al., 2009
drocharitaceae Hydrocharis morsus-ranae 1 Schoelynck et al., 2009
Haloragaceae Myriophyllum spicatum 2.1 Schoelynck et al., 2009
Nymphaeaceae Nuphar lutea 0.8 Schoelynck et al., 2009
Potamogetonaceae Potamogeton cripus 0.9 Schoelynck et al., 2009
Potamogetonaceae Potamogeton lucens 1.4 Schoelynck et al., 2009
Potamogetonaceae Potamogeton natans 1 Schoelynck et al., 2009
Potamogetonaceae Potamogeton perfoliatus 1.1 Schoelynck et al., 2009
Alismataceae Sagittaria saittifolia 2.3 Schoelynck et al., 2009
Poaceae Sparganium emersum 1.35 Schoelynck et al., 2009
Apiaceae Berula erecta 0.4 Schoelynck et al., 2009
Poaceae Glyceria maxima 1.3 Schoelynck et al., 2009
Lamiacae Mentha aquatica 0.3 Schoelynck et al., 2009
Poaceae Phalaris arundinacea 1.3 Schoelynck et al., 2009
Poaceae Phargmites australis 1.5 Schoelynck et al., 2009
Brassicaceae Rorippa amphibia 0.4 Schoelynck et al., 2009
Caprifoliceae Symphoricapos occidentalis 0.105 Bezeau et al. 1966
Fabaceae Vicia americana 0.041 Bezeau et al. 1966
Ericaceae Lyonia ferruginea 0.029 Kalisz et al., 1984
Poaceae Achnatherum hymenoides 2 Blank et al. 1994
Poaceae Elymus elymoides 1.6 Blank et al. 1994
Poaceae Cortaderia selloana 0.65 Ma and Takahashi., 2002
Poaceae Cymbopogon citratus 0.85 Ma and Takahashi., 2002
Poaceae Saccharum officinarum 0.77 Ma and Takahashi., 2002
Poaceae Danthonia intermedia 3.08 Bezeau et al. 1966
Poaceae Elymus cinereus 2.1 Bezeau et al. 1966
Poaceae Elymus innovatus 2.13 Bezeau et al. 1966
Poaceae Festuca idahoensis 3.59 Bezeau et al. 1966
Page 94
81
Poaceae Bromus inermis 2.47 Bezeau et al. 1966
Poaceae Festuca rubra 2.82 Bezeau et al. 1966
Poaceae Phleum pratense 1.59 Bezeau et al. 1966
Poaceae Andropogon gerardi 5.66 Geis, 1978.
Poaceae Sorgastrum nutans 9.41 Geis, 1978
Poaceae Saccharum officinarum 1.91 Lanning and Eleuterius, 1985
Poaceae Sporobolus cryptandrus 3.5 Smith et al., 1971
Poaceae Boutelous curtipendula 5.64 Smith et al., 1971
Poaceae Festuca scabrella 2.79 Johnston et al., 1966
Poaceae Festuca scabrella 2.19 Johnston et al., 1966
Poaceae Festuca scabrella 3.41 Johnston et al., 1966
Poaceae Festuca scabrella 3.61 Johnston et al., 1966
Poaceae Festuca scabrella 3.9 Johnston et al., 1966
Poaceae Festuca scabrella 2.4 Johnston et al., 1966
Poaceae Festuca scabrella 2.94 Johnston et al., 1966
Poaceae Festuca scabrella 4.72 Johnston et al., 1966
Poaceae Festuca scabrella 4.33 Johnston et al., 1966
Poaceae Festuca scabrella 3.35 Johnston et al., 1966
Poaceae Festuca scabrella 3.39 Johnston et al., 1966
Poaceae Festuca scabrella 2.39 Johnston et al., 1966
Poaceae Festuca scabrella 3.73 Johnston et al., 1966
Poaceae Bromus inermis 1.3 Robbins et al., 1987
Poaceae Bromus inermis 0.8 Robbins et al., 1987
AVERAGE
1.97
Page 95
82
Table of biogenic silica content in American coniferous tree species as % dry weight.
Family Genus Species %Bsi Reference
Pinaceae Abies alba 0.006 Carnelli et al., 2001
Pinaceae Juniperus nana 0.007 Carnelli et al., 2001
Pinaceae Larix decidua 0.109 Carnelli et al., 2001
Pinaceae Picea abies 0.085 Carnelli et al., 2001
Pinaceae Pinus cembra 0.011 Carnelli et al., 2001
Pinaceae Pinus mugo 0.009 Carnelli et al., 2001
Pinaceae Pinus strobus 0.085 Klein and Geis, 1978
Pinaceae Pinus resinosa 0.083 Klein and Geis, 1978
Pinaceae Pinus banksiana 0.184 Klein and Geis, 1978
Pinaceae Pinus sylvestris 0.183 Klein and Geis, 1978
Pinaceae Larix laricina 0.237 Klein and Geis, 1978
Pinaceae Larix decidua 1.372 Klein and Geis, 1978
Pinaceae Picea rubens 0.434 Klein and Geis, 1978
Pinaceae Picea mariana 0.168 Klein and Geis, 1978
Pinaceae Picea glauca 1.048 Klein and Geis, 1978
Pinaceae Pseudotsuga menziesii 0.289 Klein and Geis, 1978
Pinaceae Pseudotsuga menziesii 0.453 Klein and Geis, 1978
Pinaceae Tsuga canadensis 0.162 Klein and Geis, 1978
Pinaceae Tsuga caroliniana 0.138 Klein and Geis, 1978
Pinaceae Abies balsamea 0.182 Klein and Geis, 1978
Pinaceae Abies fraseri 0.129 Klein and Geis, 1978
Pinaceae Larix leptolepis 4.08 Ovington, 1956
Pinaceae Pinus nigra 2.82 Ovington, 1956
Pinaceae Pseudotsuga taxifolia 5.1 Ovington, 1956
Pinaceae Larix eurolepis 4.33 Ovington, 1956
Pinaceae Thuja eurolepis 3.97 Ovington, 1956
Pinaceae Pseudotsuga menziesii 1.13 Cornelis et al., 2010.
Pinaceae Picea abies 0.97 Cornelis et al., 2010.
Pinaceae Pinus negra 0.05 Cornelis et al., 2010.
Pinaceae Picea spp. 2.43 Hodson and Sangster, 1999
Pinaceae Larix spp. 2.21 Hodson and Sangster, 1999
Pinaceae Cupressocyparis spp. 1.73 Hodson and Sangster, 1999
Pinaceae Chamaecyparis spp. 1.24 Hodson and Sangster, 1999
Pinaceae Pseudotsuga spp. 0.99 Hodson and Sangster, 1999
Pinaceae Araucaria spp. 0.85 Hodson and Sangster, 1999
Pinaceae Abies spp. 0.57 Hodson and Sangster, 1999
Pinaceae Tsuga spp. 0.43 Hodson and Sangster, 1999
Pinaceae Sequoiadendron spp. 0.39 Hodson and Sangster, 1999
Pinaceae Taxus spp. 0.28 Hodson and Sangster, 1999
Pinaceae Pinus spp. 0.2 Hodson and Sangster, 1999
Page 96
83
Pinaceae Cryptomeria spp. 0.18 Hodson and Sangster, 1999
Pinaceae Cedrus spp. 0.09 Hodson and Sangster, 1999
Pinaceae Taxodium spp. 0.08 Hodson and Sangster, 1999
Pinaceae Cunninghamia spp. 0.06 Hodson and Sangster, 1999
Pinaceae Pseudolarix spp. 0.06 Hodson and Sangster, 1999
Pinaceae Thuja spp. 0.06 Hodson and Sangster, 1999
Pinaceae Juniperus spp. 0.04 Hodson and Sangster, 1999
AVERAGE
0.844
Page 97
84
Table of biogenic silica content in American deciduous tree species as % dry weight.
Family Genus Species %Bsi Reference
Salicaceae Populus tremuloides 0.011 Bezeau et al. 1966
Rosaceae Rosa woodsii 0.051 Bezeau et al. 1966
Fagaceae Quercus geminata 0.133 Kalisz et al., 1984
Betulaceae Alnus incana 0.797 Ovington, 1956
Betulaceae Betula alba 0.611 Ovington, 1956
Fagaceae Quercus spp 0.424 Ovington, 1956
Fagaceae Quercus peiraea 0.399 Ovington, 1956
Nothofagaceae Nosbofagus obliqua 0.467 Ovington, 1956
Fagaceae Quercus rubra 0.414 Ovington, 1956
Fagaceae Fagus sylvatica 0.159 Cornelis et al., 2010
Fagaceae Quercus spp. 0.117 Cornelis et al., 2010
Salicaceae Salix spp. 0.05 Bezeau et al. 1966
Fagaceae Quercus myrtifolia 0.044 Kalisz et al., 1984
Fagaceae Quercus chapmanii 0.038 Kalisz et al., 1984
Fagaceae Quercus geminata 0.133 Kalisz et al., 1984
Betulaceae Alnus viridis 0.13 Carnelli et al., 2001
Saliceae Salix helvetica 0.06 Carnelli et al., 2001
Salicaceae Populus deltides 0.94 Geis, 1978
Juglandaceae Julgans nigra 0.28 Geis, 1978
Juglandaceae Julgans cinerea 0.48 Geis, 1978
Juglandaceae Carya cordiformis 0.26 Geis, 1978
Juglandaceae Carya ovata 0.46 Geis, 1978
Juglandaceae Carya laciniosa 0.32 Geis, 1978
Juglandaceae Carya tomentosa 0.36 Geis, 1978
Betulaceae Carpinus caroliniana 0.62 Geis, 1978
Betulaceae Ostrya virginiana 0.31 Geis, 1978
Fagaceae Quercus rubra 0.44 Geis, 1978
Fagaceae Quercus alba 0.9 Geis, 1978
Fagaceae Quercus velutina 0.13 Geis, 1978
Fagaceae Quercus imbricaria 0.38 Geis, 1978
Fagaceae Quercus marcrocarpa 0.44 Geis, 1978
Fagaceae Quercus muehlenbergii 0.58 Geis, 1978
Ulmaceae Ulmus americana 3.3 Geis, 1978
Ulmaceae Ulmus rubra 2.14 Geis, 1978
Moraceae Maclura pomifera 1.2 Geis, 1978
Moraceae Morus rubra 3.79 Geis, 1978
Lauraceae lindera benzoin 0.08 Geis, 1978
Lauraceae Sassafras albidum 0.07 Geis, 1978
Platanaceae Plantanus occidentalis 0.42 Geis, 1978
Rosaceae Prunus serolina 0.04 Geis, 1978
Page 98
85
Rosaceae Prunus viginiana 0.52 Geis, 1978
Aceraceae Acer saccharum 0.98 Geis, 1978
Aceraceae Acer negundo 0.34 Geis, 1978
Tiliaceae Tilica americana 0.49 Geis, 1978
Cornaceae Cornus stolonifera 0.2 Geis, 1978
Oleaceae Fraxinus americana 0.42 Geis, 1978
AVERAGE
0.54
Page 99
86
Appendix II.
Calculated terrestrial BSi model parameters for each catchment
Land Cover Gauges
Average
Annual
DSi
Average
Annual
Precip
Drainage
Area
Average
Annual
Discharge
DSi Flux DBSi Flux
NPP %
BSi
BSi Annual
Production BSi Dissolution
Oct to Apr NPP · % BSi (SW·Q)/(M · SAA)
mg/l in ha f3/s kg/ha · yr kg/ha · yr kg/ha · yr kg/ha · yr kg/ha · yr
Grasslands
8064700 9.65 36.06 50762 117 2.14 - 3887.21 2.30 89.45 3.29
5451210 19.13 41.69 58013 179 5.55 3.79 2984.91 2.30 68.69 6.52
5451080 24.67 25.10 20719 18 1.99 1.51 3418.87 2.30 78.67 8.41
6306200 10.79 15.45 22868 29 1.21 0.92 2173.58 2.30 50.02 3.68
6306200 11.25 15.45 22868 18 0.79 - 2533.77 2.30 58.31 3.84
6295113 14.20 18.45 31855 14 0.64 - 2334.95 2.30 53.73 4.84
6340000 8.64 23.25 116281 56 0.37 - 2731.76 2.30 62.86 2.95
6342260 10.59 26.47 37812 35 0.77 - 2720.28 2.30 62.60 3.61
5056340 15.31 20.94 82876 44 0.80 - 2892.45 2.30 66.56 5.22
Wetlands
1022890 2.38 44.91 510 2 0.99 0.16 10754.87 1.97 212.03 8.11
2310947 6.13 45.46 72517 83 0.62 0.72 6581.81 1.97 129.76 20.91
2299950 7.39 42.75 16912 36 1.50 0.81 6182.81 1.97 121.89 25.20
Coniferous
Forest
11264500 6.23 46.95 46877 403 5.11 3.00 902.17 0.84 7.62 20.22
10343500 26.78 37.92 2719 10 9.26 7.18 1534.89 0.84 12.97 86.92
5014300 2.31 102.67 3755 109 6.04 2.50 730.00 0.84 6.17 7.50
9196500 2.05 21.77 19631 168 1.59 0.96 626.70 0.84 5.30 6.65
5124480 1.50 20.00 65783 166 0.30 - 1404.91 0.84 9.87 4.87
6102500 4.60 19.00 28488 176 2.55 - 1004.91 0.84 11.87 14.93
6279795 24.00 23.00 4920 41 17.98 - 1326.91 0.84 118.88 77.90
6332515 7.60 29.00 19165 3 5.35 - 936.91 0.84 7.24 24.67
Deciduous
Forest
1349950 1.50 45.13 17766 154 1.25 0.97 2382.83 0.54 12.91 6.53
1362380 1.10 56.17 8158 122 1.45 0.69 2245.09 0.54 12.17 4.79
1545600 1.78 43.49 11965 72 0.96 0.62 2751.77 0.54 14.91 7.75
4063700 4.22 20.22 35999 72 0.80 0.49 2666.91 0.54 14.45 18.41
1364959 0.96 56.78 1388 21 1.21 0.73 2589.43 0.54 14.03 4.18
1422747 1.66 45.13 6397 56 1.34 0.90 2088.42 0.54 11.32 7.22
Page 100
87
Calculated terrestrial BSi model parameters for each catchment, continued
from above
Gauges
BSi Storage DBSi Storage Leaching
Rate BSi Storage DBSi Storage
BSipro -
BSidis
BSidis -
BSistream R
2-value
BSipro -
BSidis
BSidis -
BSistream
kg/ha · yr kg/ha · yr kg/ha · yr kg/ha · yr kg/ha · yr
Grasslands
8064700 86.16 1.15 1.22 86.16 1.15
5451210 62.17 0.97 3.16 62.17 0.97
5451080 70.26 6.43 1.13 70.26 6.43
6306200 46.34 2.47 0.69 46.34 2.47
6306200 54.47 3.04 0.45 54.47 3.04
6295113 48.89 4.2 0.36 48.89 4.2
6340000 59.92 2.58 0.21 59.92 2.58
6342260 58.99 2.84 0.44 58.99 2.84
5056340 61.34 4.42 0.46 61.34 4.42
Wetlands
1022890 203.93 7.86 0.25 203.93 7.86
2310947 108.85 20.76 0.15 108.85 20.76
2299950 96.7 24.82 0.37 96.7 24.82
Coniferous Forest
11264500 -12.6 16.23 3.99 -12.6 16.23
10343500 -73.95 79.7 7.22 -73.95 79.7
5014300 -1.33 2.79 4.71 -1.33 2.79
9196500 -1.36 5.41 1.24 -1.36 5.41
5124480 7 4.63 0.23 7 4.63
6102500 -3.06 12.94 1.99 -3.06 12.94
6279795 -66.03 63.88 14.02 -66.03 63.88
6332515 -12.8 20.5 4.17 -12.8 20.5
Deciduous Forest
1349950 6.39 6.11 0.41 6.39 6.11
1362380 7.38 4.31 0.48 7.38 4.31
1545600 7.16 7.44 0.32 7.16 7.44
4063700 -3.96 18.15 0.26 -3.96 18.15
1364959 9.85 3.79 0.4 9.85 3.79
1422747 4.1 6.78 0.44 4.1 6.78