Additional File 8

Data Extraction Proforma and Strategy

Table A7.1. Tabulated layout for data extraction spreadsheet.

Variable / Data
Author
Year
Title
Journal
Region
Country
Soil description
Intervention/ Exposure
Comparator
Intervention Category
Outcomes
Design
Method
Data location in paper
Data format
Reported Units
Mean comparator
Mean intervention
SD comparator
SD intervention
n comparator
n intervention
Comments

Data Extraction Strategy

The following illustrate the key decisions made when presented with choices in data extraction.

  1. BACI (before-after-comparator-intervention) study design data were converted to change-over-time for intervention and comparator data by subtracting baselines from the latest point of measurement. An effect size based on difference between the intervention and comparator could then be calculated.
  2. Where multiple years following an intervention were reported, data from the latest time point were used as ‘after’ data, since land management changes are likely to act over long periods of time. Data collected from within 12 months of the intervention was excluded where later data was available. In one study where ‘after’ data appeared highly variable, data from multiple years was pooled so as to be more conservative.
  3. Where studies reported data from geographically separated sites or sites that were known to differ substantially in ecology, these data were treated as different study units, provided that both comparator and intervention data were present for each site. Hence, the unit of measurement in the synthesis was the study site. Where measures of variability were not available for single sites, standard deviations were calculated across sites using site means.
  4. Where spatial replicates were measured on more than one occasion (for example flux chambers measured over a year) the longest possible reporting period was used to gather a pooled mean and standard deviation.
  5. Although technically pseudoreplication, the majority of studies did not replicate at the intervention/exposure level. Therefore, so as to retain sufficient studies any form of spatial replication was treated as an independent unit of replication. Sample size, therefore, was taken to be the number of samples taken across space, not time. Purely temporal replication (i.e. repeated measurements of the same sample) was excluded from consideration. This was decided upon since such studies violate assumptions of independence of data. Furthermore, the results of models using temporal variability can only be generalised across time and not space. Spatial variability reported for multiple time periods was pooled (pooled standard deviation). Where temporal and spatial variability could not be discerned, data were included in the meta-analysis but sensitivity analysis was carried out to determine the implication of doing so. Pooled standard deviations were calculated according to the equation;

SDpooled=n1-1S12+n2-1S22n1+n2-2

  1. Variability was converted from confidence intervals and standard errors into the common rubric of standard deviation where sample sizes were reported. Where details of sample size were not precise (for example “more than 30 samples were taken”) a conservative estimate was used based on best information; i.e. the smaller sample size was used in further calculations. Where no measure of variability was provided, means were extracted for inclusion in narrative synthesis alone.
  2. Where data were reported over within-site spatial scales that were not commonly reported enough to allow inclusion as effect modifiers across studies (for example soil depth or peat microform), means and standard deviations were pooled and reported as summary statistics for that replicate.
  3. Where multiple interventions or exposures were recorded across a gradient (for example drained and partially drained relative to an undrained control) data were extracted for the more extreme case (for example fully drained rather than partially drained). This was to ensure the fullest extent of the study results was included in the meta-analysis.
  4. Where multiple potential interventions/exposures were reported within one manuscript/study and they did not have an obvious comparator, a meaningful comparator was chosen such that there was consistency between studies concerning the same land management.
  5. For studies that examined exposures rather than interventions, the comparator was chosen as the more natural state, for example rich peat quality, high water table, or wet soil conditions.
  6. Where CO2 data were reported, total ecosystem respiration (Rtot) and net ecosystem exchange (NEE) were extracted. Gross ecosystem photosynthesis (GEP) was not extracted since it is calculated from NEE minus Rtot. Where GEP and Rtot alone were reported, NEE was calculated.

Table A7.2. Information regarding data location, extraction strategy and pooling calculations for means and standard deviations for studies included in the synthesis.

Article / Region / Outcomes / Quantitative data locations in paper / Data Extraction Method / Pooling / Pooled Across
Augustin, J., W. Merbach, et al. (1998) Factors influencing nitrous oxide and methane emissions from minerotrophic fens in northeast Germany.. Biology and Fertility of Soils 28(1): 1-4. / Gumnitzniederung / CH4 / Figure 1 / Data Thief / Yes / Time (mean only)
Augustin, J., W. Merbach, et al. (1998) Factors influencing nitrous oxide and methane emissions from minerotrophic fens in northeast Germany.. Biology and Fertility of Soils 28(1): 1-4. / Gumnitzniederung / N2O / Figure 1 / Data Thief / Yes / Time (mean only)
Augustin, J., W. Merbach, et al. (1998) Factors influencing nitrous oxide and methane emissions from minerotrophic fens in northeast Germany.. Biology and Fertility of Soils 28(1): 1-4. / Rhin-Havelluch / N2O / Figure 2 / Data Thief / Yes / Time (mean only)
Augustin, J., W. Merbach, et al. (1998) Nitrous oxide fluxes of disturbed minerotrophic peatlands.. Agribiological Research-Zeitschrift Fur Agrarbiologie Agrikulturchemie Okologie 51(1): 47-57. / Mucheberg (Gumnitz/Bradndenburg) / N2O / Figure 3 and Table 2 / Data Thief / Yes / Time (mean only)
Augustin, J., W. Merbach, et al. (1998) Nitrous oxide fluxes of disturbed minerotrophic peatlands.. Agribiological Research-Zeitschrift Fur Agrarbiologie Agrikulturchemie Okologie 51(1): 47-57. / Paulinenaue (Rhin-Havel-Luch/Brandenburg) / N2O / Figure 1 and Table 2 / Data Thief / Yes / Time (mean only)
Augustin, J., W. Merbach, et al. (1998) Nitrous oxide fluxes of disturbed minerotrophic peatlands.. Agribiological Research-Zeitschrift Fur Agrarbiologie Agrikulturchemie Okologie 51(1): 47-57. / Heinrichswalde (Friedlander Grosse Wiese/Mecklenburg-Volpommern) / N2O / Figure 2 and Table 2 / Direct Extraction / No
Augustin, J., W. Merbach, et al. (1998) Nitrous oxide fluxes of disturbed minerotrophic peatlands.. Agribiological Research-Zeitschrift Fur Agrarbiologie Agrikulturchemie Okologie 51(1): 47-57. / Heinrichswalde (Friedlander Grosse Wiese/Mecklenburg-Volpommern) / N2O / Table 2 / Data Thief / Yes / Time (mean only)
Banas, K. and K. Gos (2004) Effect of peat-bog reclamation on the physico-chemical characteristics of the ground water in peat.. Polish Journal of Ecology 52(1): 69-74. / Pomerania / DOC / Table 1, p 71 / Direct Extraction / Yes / Replicates
Bubier, J. L., T. R. Moore, et al. (2007) Effects of nutrient addition on vegetation and carbon cycling in an ombrotrophic bog. Global Change Biology 13(6): 1168-1186. / Mer Bleue / NEE CO2 / Figures 1 and 2 / Data Thief / Yes / Replicates
Bubier, J. L., T. R. Moore, et al. (2007) Effects of nutrient addition on vegetation and carbon cycling in an ombrotrophic bog. Global Change Biology 13(6): 1168-1186. / Mer Bleue / Rtot CO2 / Figures 1 and 2 / Data Thief / Yes / Replicates
Cagampan, J. P. and J. M. Waddington (2008) Net ecosystem CO2 exchange of a cutover peatland rehabilitated with a transplanted acrotelm.. Ecoscience 15(2): 258-267. / Pointe-Lebel / NEE CO2 / Figure 5 / Data Thief / No
Cagampan, J. P. and J. M. Waddington (2008) Net ecosystem CO2 exchange of a cutover peatland rehabilitated with a transplanted acrotelm.. Ecoscience 15(2): 258-267. / Pointe-Lebel / Rtot CO2 / Figure 5 / Data Thief / No
Cagampan, J. P. and J. M. Waddington (2008) Net ecosystem CO2 exchange of a cutover peatland rehabilitated with a transplanted acrotelm.. Ecoscience 15(2): 258-267. / Pointe-Lebel / NEE CO2 / Figure 5 / Data Thief / No
Cagampan, J. P. and J. M. Waddington (2008) Net ecosystem CO2 exchange of a cutover peatland rehabilitated with a transplanted acrotelm.. Ecoscience 15(2): 258-267. / Pointe-Lebel / Rtot CO2 / Figure 5 / Data Thief / No
Danevcic, T., I. Mandic-Mulec, et al. (2010) Emissions of CO(2), CH(4) and N(2)O from Southern European peatlands.. Soil Biology & Biochemistry 42(9): 1437-1446. / Ljubljana Marsh / CH4 / Figure 1 and Table 2 / Direct Extraction / No
Danevcic, T., I. Mandic-Mulec, et al. (2010) Emissions of CO(2), CH(4) and N(2)O from Southern European peatlands.. Soil Biology & Biochemistry 42(9): 1437-1446. / Ljubljana Marsh / N2O / Figure 1 and Table 2 / Direct Extraction / No
Danevcic, T., I. Mandic-Mulec, et al. (2010) Emissions of CO(2), CH(4) and N(2)O from Southern European peatlands.. Soil Biology & Biochemistry 42(9): 1437-1446. / Ljubljana Marsh / Rtot CO2 / Figure 1 and Table 2 / Direct Extraction / No
Dias, A. T. C., B. Hoorens, et al. (2010) Plant Species Composition Can Be Used as a Proxy to Predict Methane Emissions in Peatland Ecosystems After Land-Use Changes.. Ecosystems 13(4): 526-538. / Nieuwkoopse Plassen / CH4 / Table 1 / Direct Extraction / Yes / Replicates
Dias, A. T. C., B. Hoorens, et al. (2010) Plant Species Composition Can Be Used as a Proxy to Predict Methane Emissions in Peatland Ecosystems After Land-Use Changes.. Ecosystems 13(4): 526-538. / Nieuwkoopse Plassen / CH4 / Table 1 / Direct Extraction / Yes / Replicates
Dias, A. T. C., B. Hoorens, et al. (2010) Plant Species Composition Can Be Used as a Proxy to Predict Methane Emissions in Peatland Ecosystems After Land-Use Changes.. Ecosystems 13(4): 526-538. / Het Guisveld / CH4 / Table 1 / Direct Extraction / Yes / Replicates
Dirks, B. O. M., A. Hensen, et al. (2000) Effect of drainage on CO2 exchange patterns in an intensively managed peat pasture.. Climate Research 14(1): 57-63. / Zegveld / NEE CO2 / Table 2 / Direct Extraction / No
Dirks, B. O. M., A. Hensen, et al. (2000) Effect of drainage on CO2 exchange patterns in an intensively managed peat pasture.. Climate Research 14(1): 57-63. / Zegveld / Rtot CO2 / Table 2 / Direct Extraction / No
Fiedler, S., B. S. Hoell, et al. (2008) Particulate organic carbon (POC) in relation to other pore water carbon fractions in drained and rewetted fens in Southern Germany.. Biogeosciences 5(6): 1615-1623. / Donauriad fen / DOC / Table 1 / Direct Extraction / Yes / Soil depth
Fiedler, S., B. S. Hoell, et al. (2008) Particulate organic carbon (POC) in relation to other pore water carbon fractions in drained and rewetted fens in Southern Germany.. Biogeosciences 5(6): 1615-1623. / Donauriad fen / POC (TOC-DOC) / Table 1 / Direct Extraction / Yes / Soil depth
Fiedler, S., B. S. Hoell, et al. (2008) Particulate organic carbon (POC) in relation to other pore water carbon fractions in drained and rewetted fens in Southern Germany.. Biogeosciences 5(6): 1615-1623. / Donauriad fen / DOC / Table 1 / Direct Extraction / Yes / Soil depth
Fiedler, S., B. S. Hoell, et al. (2008) Particulate organic carbon (POC) in relation to other pore water carbon fractions in drained and rewetted fens in Southern Germany.. Biogeosciences 5(6): 1615-1623. / Donauriad fen / POC (TOC-DOC) / Table 1 / Direct Extraction / Yes / Soil depth
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / DOC / Figure 1 / Data Thief / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / Rtot CO2 / Table 1 / Direct Extraction / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / DOC / Figure 1 / Data Thief / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / DOC / Figure 1 / Data Thief / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / DOC / Figure 1 / Data Thief / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / Rtot CO2 / Table 1 / Direct Extraction / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / Rtot CO2 / Table 1 / Direct Extraction / No
Glatzel, S., K. Kalbitz, et al. (2003) Dissolved organic matter properties and their relationship to carbon dioxide efflux from restored peat bogs.. Geoderma 113(3-4): 397-411. / Riviere de Loup / Rtot CO2 / Table 1 / Direct Extraction / No
Glenn, S., A. Heyes, et al. (1993) CARBON-DIOXIDE AND METHANE FLUXES FROM DRAINED PEAT SOILS, SOUTHERN QUEBEC.. Global Biogeochemical Cycles 7(2): 247-257. / Napierville / CH4 / Figures 1 and 2 and Table 2 / Direct Extraction / No