Anthropogenic impact on biogenic Si pools in temperate soils

Introduction Conclusions References


C2212
an influence on biological Si-cycling and Si-storage in temperate ecosystems (reviews of Sommer et al., 2006 andCornelis et al., 2011 for a more extensive list).Our broad definition of the temperate region, and the resolution of the available land use data (i.e. per country) implies that climatology, geology and pedology can't be considered spatially uniform.This will be thoroughly emphasized throughout the manuscript Total land-ocean fluxes of DSi were estimated based on the annual loss of Psia in the soil.This implies that all BSi has been dissolved and leached out of the soil, but this is a simplification of the processes present.Other processes can be regarded responsible for losses in BSi; (re)-precipitation of dissolved BSi as secondary clay minerals or less dissolvable pedogenic silica fractions (Lucas et al., 1993;Van Cappellen, 2003), direct uptake in the agricultural cycle by annual harvest, timber logging, hay-meadow and grazing (Vandevenne et al., 2011), losses through severe land degradation (Smis et al., 2010;Triplett, 2008) and soil erosion as well as sequestration within rivers & lakes (Humborg et al., 2002(Humborg et al., & 2005;;Schelske et al., 1983).Furthermore the aqueous chemistry of Si is regulated by a number of linked processes (overview Cornelis et al., 2011).The fact that not all "disappearing" BSi is exported as DSi, will be clearly mentioned in the revised version and the implications of this will be discussed.
We want to stress that the purpose of our up-scaling exercise was meant to illustrate the possible importance of reduced Si-storage on the land-ocean Si flux and the order of magnitude of the impact on Si dynamics.We delivered a first estimate based on available data.This is also appreciated and acknowledged by both reviewers in their comments.Because we can't make a more accurate estimation, we will tone down the conclusions with respect to the effect on land-ocean fluxes on temperate and global scale.We are convinced that the magnitude is representative for how much BSi has been lost and/or converted in temperate soils, and that this illustrates the importance it could have on Si-delivery to mainly aquatic systems.(et al., 2008) for temperate regions could result in an average annual net increase of 0.16 Tmol Si y-1.This value corresponds with 12% of the temperate load and 3% of the global load.The value retrieved from our Swedish study suggests significantly higher annual net losses of ca; 1.1Tmol Si yr-1 corresponding on temperate and global scale with 84% of the temperate load and 20% of the global load respectively.The large discrepancy can be explained by several factors (1) a smaller Psia pool present in the HBEF catchment (Saccone et al., 2007) (2) losses in HBEF are based on riverine measurements of DSi so precipitation and biomineralization are taken into account but particulate BSi transport is neglected and (3) perturbations in HBEF were only persistent for 3 years, after which regeneration was allowed.We now will discuss these important differences in the dynamics of Psia within the temperate region, as well as the processes that may induce differences between changes in BSi stocks and BSi delivery to rivers such as (re-)precipitation and bio-mineralisation etc.Our study indicates that large uncertainties exist mainly due to lack of studies concerning this topic, and we encourage further research towards the effect of human influences on bio-reactive Si pools.Hereby we suggest to take in account changes in dissolved silica fluxes, biogenic silica fluxes and changes in distribution and total silica pools in soils to disentangle the effect of human influence on terrestrial Si cycle and land-ocean flux.
Comment 5 R2: "Is it an increase of BSi exported and/or BSi dissolution, or is it a change of Si form (BSi dissolution and Si precipitation with other elements as secondary solid phases)?"The reviewer indeed makes an appropriate comment about the fact that it is difficult to decipher the processes responsible for the redistribution.In our manuscript we mention all different processes like preservation and physical transport, secondary mineral formation, Si adsorption and precipitation of pedogenic opal.We specifically target changes due to reduced input of BSi from vegetation in the soil as the plant-soil cycle is perturbed, and the mobilisation of BSi within the soil profile (e.g.leaching, physical transport and (re)-precipitation.These are important drivers for long-term losses.Transformation to other Si forms will play also a role, but C2214 this correlates with a perturbation of the soil-vegetation continuum.A complex linkage of different processes, which needs to be investigated further, is beyond the scope the paper.Still we will clarify the presence of other controlling factors or processes to make to reader aware of the complexity.
Comment 6 R2: "alkali-extractable Si" or BSi, please standardize in your manuscript or redefine" We will adapt the BSi term, because it does not only constitute phytoliths and microscope plant residues (i.e.real biogenic fraction, see Sauer et al., 2006), but also a poorly-crystalline fraction and amorphous lithogenic and pedogenic forms (e.g.silica glass, silica sorbed on Fe and Al oxi/hydroxides).It could be assumed that the importance of poorly-crystalline fraction in the total pool can be neglected.The synthesis of allophane and imogolite fractions, mainly occurring in volcanic soils, is hampered due to the acidic conditions (pH: 3.3-4.7).Although other alkali-extracted silica fractions are not present in a biogenic form, they have generally a higher solubility then the mineral phase.This makes them more easily available for biological (re-)cycling on shorter time scales (Sommer et al., 2006).Moreover Si adsorbed on Fe-and Al oxi-/hydroxides can be cycled by plants before sorption.This fraction partly controls the concentration of H4SiO4 in aqueous phase (Cornelis et al., 2011).If leached, it would be replenished by Si from phytoliths (Farmer et al., 2005).As the alkali-extracted silica here mainly corresponds with the amorphous silica fraction, we will standardise the alkali-extracted silica with the term amorphous silica (ASi) throughout the manuscript.
Comment 7 R2: "I don't agree with your definition of the Si extracted by CaCl2."We will rephrase our definition of this easily soluble silica to plant-available silica without referring to phytolith origin.However, as hypothesized by Farmer et al. (2005), we do consider phytoliths as the main Si source controlling the final equilibrium concentrations of DSi in soil water.We agree that the control of phytoliths on the silica concentration in the aqueous phase depends on climatology and weathering stage (Cornelis et al., 2011), indicating that in non-weathered soils adsorbed silica also plays in important role.For our study site the lower CSie concentrations, and associated lower

C2215
Psie, indicate that the magnitude of BSi dynamics is less under arable land then under other sites, so that a lower Si-equilibrium stage is reached.We will change/rephrase our conclusions, and will indicate that we hypothesize that this mainly represents only dissolution of phytoliths, acknowledging other potential pathways.
Comment 8 R2: "the OC content is a proxy of the rate of organic matter input in topsoil and is not a direct chemical proxy of the BSi presence or absence."We agree with the comment of reviewer 2. OC content can't be regarded as a proxy for the presence or absence of BSi and we will not longer use the term 'proxy' in the revised manuscript.However, it makes sense to investigate how OC content and CSIa are related because we know that litter fall, and decomposition is an important source of both organic matter and biogenic silica, e.g.phytoliths, to the top-layer.We will argue in the discussion two reasons exist why this can't be true for deeper layers: (1) there is no strict coupling between OM and biogenic silica and (2) different processes are responsible for the redistribution of OC and CSia.This implicitly means that once pedogenic processes are involved OC can't be regarded as a proxy nor predictor.We will rewrite the paragraph to clarify this misunderstanding.
Comment 9 R2: "phytolith from roots decomposition?Other studies in specific plant species shown a very low content of Si in roots (see for instance Gérard et al. 2008)." In the present estimates, only the contribution of the above-ground biomass is considered due to the lack of data on below-ground productivity.This omission will inevitably lead to an underestimation of the total input.We wanted to address in our paper the possibility of extra input at various depths.Little is known about concentrations in roots and their turnover rate.Turnover rates have shown to be significantly different, between 10-56% annually, for roots in different vegetation types (e.g.trees, shrub, grassland) and for different root sizes (Gill & Jackson, 2000).Gérard et al. (2008) refers to a value measured in tree roots by Gordon & Jackson (2000), which we however couldn't retrieve in this original paper.Still, Webb and Longstaffe (2000) estimated that roots and rhizomes contained up to 34% of the silica content of two grass species.Another paper C2216 on grass species showed that the amount of opal deposited in the soil annually by root systems and above-ground parts is approximately equal in magnitude (Geis, 1977).An older review paper on 'Silicon and plant growth' (Lewin & Reimann, 1969) states that dependent on plant type, silicon can be uniformly distributed between shoots and roots, or accumulate in shoots or roots preferentially.Accurate measurements of root Si input for different plant species and for larger scale ecosystems are lacking.We will include the root discussion because we want to clarify different processes which can be regarded responsible for difference in distribution.In current state-of-the-art, as shown above, roots cannot be excluded.
Technical comments: We will address all smaller concerns of both reviewers as well as technical comments (rephrasing, references, figure adaptions and smaller clarifications) in the final revised version.
Interactive comment on Biogeosciences Discuss., 8, 4391, 2011. C2217 Comment4 R1: "I suggest to apply the land-use change related increase rate of ï ň Ćuvial DSi ï ň Ćuxes from Conley et al. (2008)" A simplified calculation of excess silica loss based on the rate of Conley