Topsoil nutrient data - total nitrogen (N) concentration (%), C:N ratio and Olsen-Phosphorus (mg/kg). Data is representative of 0 - 15 cm soil depth. Cores from 256 1km x 1km squares across Great Britain were analysed in 2007. For total N concentration (and therefore C:N ratio), a total of 1024 cores were analysed, and for Olsen-P, a total of 1054 cores were analysed. See Emmett et al. 2010 for further details of sampling and methods (http://nora.nerc.ac.uk/id/eprint/5201/1/CS_UK_2007_TR3%5B1%5D.pdf) Estimates of mean values within selected habitats and parent material characteristics across GB were made using Countryside Survey (CS) data from 1978, 1998 and 2007 using a mixed model approach. The estimated means of habitat/parent material combinations are modelled on dominant habitat and parent material characteristics derived from the Land Cover Map 2007 and Parent Material Model 2009, respectively. The parent material characteristic used was that which minimised AIC in each model (see Dataset Documentation). Please see Scott, 2008 for further details of similar statistical analysis (http://nora.nerc.ac.uk/id/eprint/5202/1/CS_UK_2007_TR4%5B1%5D.pdf). Areas, such as urban and littoral rock, are not sampled by CS and therefore have no associated data. Also, in some circumstances sample sizes for particular habitat / parent material combinations were insufficient to estimate mean values. Full details about this dataset can be found at https://doi.org/10.5285/7055965b-7fe5-442b-902d-63193cbe001c

Countryside Survey topsoil pH and bulk density (g cm-3) data is representative of 0 - 15 cm soil depth. Topsoil pH was measured using 10g of field moist soil with 25ml de-ionised water giving a ratio of soil to water of 1:2.5 by weight; bulk density was estimated by making detailed weight measurements throughout the soil processing procedure. For topsoil pH and bulk density data, a total of 2614 cores from 591 1km x 1km squares across Great Britain were collected and analysed in 2007. Please see Emmett et al. 2010 for further details of sampling and methods (http://nora.nerc.ac.uk/id/eprint/5201/1/CS_UK_2007_TR3%5B1%5D.pdf). Estimates of mean values within selected habitats and parent material characteristics across GB were made using CS data from 1978, 1998 and 2007 using a mixed model approach. Please see Scott, 2008 for further details of similar statistical analysis (http://nora.nerc.ac.uk/id/eprint/5202/1/CS_UK_2007_TR4%5B1%5D.pdf). The estimated means of habitat /parent material combinations using 2007 data are modelled on dominant habitat and parent material characteristics derived from the Land Cover Map 2007 and Parent Material Model 2009, respectively. The parent material characteristic used was that which minimised AIC in each model (see Dataset Documentation). The Countryside Survey looks at a range of physical, chemical and biological properties of the topsoil from a representative sample of habitats across the UK. Countryside Survey soils data are freely available under licence from the Environmental Information Data Centre catalogue. Full details about this dataset can be found at https://doi.org/10.5285/5dd624a9-55c9-4cc0-b366-d335991073c7

Vegetation map of Plynlimon catchments. The vegetation map was created for assisting with identifying soil survey vegetation boundaries, and is not an authoritative vegetation survey. The defined vegetation classes correspond with the classes defined in CEH Land Cover Map 2000 (LCM 2000). LCM2000 was classified using a hierarchical nomenclature corresponding to the Joint Nature Conservation Committee (JNCC) Broad Habitats which encompasses the entire range of UK habitats.

Topsoil microbe data - Bacteria. Data is representative of 0 - 15 cm soil depth and includes bacterial community structure as assessed by ordination scores, Shannon diversity index and Simpson diversity index. Over 1000 samples from 233 1km x 1km squares across Great Britain were sampled in 2007 and analysed. See Emmett et al. 2010 for further details of sampling and Griffiths et al. 2011 for further details of methods. Estimates of mean values within selected habitats and parent material characteristics across GB were made using Countryside Survey (CS) data from 2007 using a mixed model approach. The estimated means of habitat/parent material combinations using 2007 data are modelled on dominant habitat and parent material characteristics derived from the Land Cover Map 2007 and Parent Material Model 2009, respectively. The parent material characteristic used was that which minimised AIC in each model (see dataset documentation). Please see Scott, 2008 for further details of similar statistical analysis. Areas, such as urban and littoral rock, are not sampled by CS and therefore have no associated data. Also, in some circumstances sample sizes for particular habitat/parent material combinations were insufficient to estimate mean values. References: Emmett, B.A., Frogbrook, Z.L., Chamberlain P.M., Griffiths R., Pickup R., Poskitt, J., Reynolds B., Rowe E., Rowland P., Spurgeon D., Wilson J., Wood, C.M. (2008). Countryside Survey Technical Report No.03/07: Soils Manual. Centre for Ecology & Hydrology. http://nora.nerc.ac.uk/id/eprint/5201/1/CS_UK_2007_TR3%5B1%5D.pdf Griffiths, R.I.; Thomson, B.C.; James, P.; Bell, T.; Bailey, M.; Whiteley, A.S. (2011). The bacterial biogeography of British soils. Environmental Microbiology, 13 (6). 1642-1654. https://doi.org/10.1111/j.1462-2920.2011.02480.x Scott, W.A. (2008). CS Technical Report No.4/07: Statistical Report. Centre for Ecology & Hydrology. http://nora.nerc.ac.uk/id/eprint/5202/1/CS_UK_2007_TR4%5B1%5D.pdf Full details about this dataset can be found at https://doi.org/10.5285/53210c27-87fc-46e4-a3d6-e731003dc541

Topsoil moisture data - gravimetric moisture content (%). Data is representative of 0 - 15 cm soil depth. Cores from 1098 plots within 256 1km by 1km squares were measured in 1998 and 2614 plots within 591 1km x 1km squares were analysed in 2007 across Great Britain. See Emmett et al. 2010 for further details of sampling and methods (http://nora.nerc.ac.uk/id/eprint/5201/1/CS_UK_2007_TR3%5B1%5D.pdf). Estimates of mean values within selected habitats and parent material characteristics across GB were made using Countryside Survey (CS) data from 1998 and 2007 using a mixed model approach. The estimated means of habitat/parent material combinations are modelled on dominant habitat and parent material characteristics derived from the Land Cover Map 2007 and Parent Material Model 2009, respectively. The parent material characteristics used were those which minimised AIC in the model (see Data documentation). Please see Scott, 2008 for further details of similar statistical analysis (http://nora.nerc.ac.uk/id/eprint/5202/1/CS_UK_2007_TR4%5B1%5D.pdf). Areas, such as urban and littoral rock, are not sampled by CS and therefore have no associated data. Also, in some circumstances sample sizes for particular habitat / parent material combinations were insufficient to estimate mean values. Full details about this dataset can be found at https://doi.org/10.5285/8db84900-5fdb-43be-a607-e56c843d9b87

Countryside Survey topsoil invertebrate data is representative of 0 - 8 cm soil depth and includes Total catch, Mite:Springtail ratio, Number of broad taxa and Shannon diversity. For invertebrate data, a total of 947 cores from 256 1km x 1km squares across Great Britain were analysed in 2007. Please see Emmett et al. 2010 for further details of sampling and methods (http://nora.nerc.ac.uk/id/eprint/5201/1/CS_UK_2007_TR3%5B1%5D.pdf ). Estimates of mean values within selected habitats and parent material characteristics across GB were made using CS data from 1978, 1998 and 2007 using a mixed model approach. Please see Scott, 2008 for further details of similar statistical analysis (http://nora.nerc.ac.uk/id/eprint/5202/1/CS_UK_2007_TR4%5B1%5D.pdf ). The estimated means of habitat /parent material combinations using 2007 data are modelled on dominant habitat and parent material characteristics derived from the Land Cover Map 2007 and Parent Material Model 2009, respectively. The parent material characteristic used was that which minimised AIC in each model (see Supporting Information). Areas, such as urban and littoral rock, are not sampled by CS and therefore have no associated data. Also, in some circumstances sample sizes for particular habitat / parent material combinations were insufficient to estimate mean values. Full details about this dataset can be found at https://doi.org/10.5285/f19de821-a436-4b28-95f6-b7287ef0bf15

These catchment boundaries define upland catchments and subcatchments at the headwaters of rivers Upper Hafren (Severn) and Upper Gwy (Wye). They identify the area of study of the Plynlimon research catchment project.

The Land Classification 2007 is a classification of Great Britain into a set of 45 environmental strata, termed land classes, to be used as a basis for ecological survey, originally developed by the Institute of Terrestrial Ecology (ITE) in the late 1970s and building upon previous 1990 and 1998 versions. The strata were created from the multivariate analysis of 75 environmental variables, including climatic data, topographic data, human geographical features and geology data. The Land Classification can be used to stratify a wide range of ecological and biogeographical surveys to improve the efficiency of collection, analysis and presentation of information derived from a sample. The Land Classification 2007 provided stratification for the Countryside Survey of Great Britain 2007 and was adjusted from the 1998 version as a consequence of needing to provide National Estimates for habitats for Wales from Countryside Survey 2007 in addition to Scotland and GB as a whole. Both the 1990 version and the 1998 versions are also available. Full details about this dataset can be found at https://doi.org/10.5285/5f0605e4-aa2a-48ab-b47c-bf5510823e8f

The Land Classification 1998 is a classification of Great Britain into a set of 40 environmental strata, termed land classes, to be used as a basis for ecological survey, originally developed by the Institute of Terrestrial Ecology (ITE) in the late 1970s and building upon a previous 1990 version. The strata were created from the multivariate analysis of 75 environmental variables, including climatic data, topographic data, human geographical features and geology data. The Land Classification can be used to stratify a wide range of ecological and biogeographical surveys to improve the efficiency of collection, analysis and presentation of information derived from a sample. The Land Classification 1998 provided stratification for the Countryside Survey of Great Britain 2000 and was adjusted from the 1990 version as a consequence of needing to provide National Estimates for habitats for Scotland from Countryside Survey 2000 in addition to GB as a whole. The dataset was later modified in 2007 during Countryside Survey 2007. Both the 1990 version and the 2007 version are also available. Full details about this dataset can be found at https://doi.org/10.5285/971671a6-98b4-4d80-b165-21dace7373b9

The Land Classification 1990 is a classification of Great Britain into a set of 32 environmental strata, termed land classes, to be used as a basis for ecological survey, originally developed by the Institute of Terrestrial Ecology (ITE) in the late 1970s. The strata were created from the multivariate analysis of 75 environmental variables, including climatic data, topographic data, human geographical features and geology data. The Land Classification can be used to stratify a wide range of ecological and biogeographical surveys to improve the efficiency of collection, analysis and presentation of information derived from a sample. The Land Classification 1990 provided stratification the Countryside Survey of Great Britain 1990. The dataset was later modified in 1998 and 2007 for successive Countryside Surveys, both versions of which are also available. Full details about this dataset can be found at https://doi.org/10.5285/ab320e08-faf5-48e1-9ec9-77a213d2907f