Alternative site for the Global ecosystem typology with additional information for ecosystem profiles and indicative maps.
This site is maintained by jrfep
This table summarizes information regarding sources used for generating the indicative maps for all Ecosystem Functional Groups. Alternatively, follow this link to see the list of references used for the maps.
| Map code | Description | Contributors |
|---|---|---|
| F1.1 Permanent upland streams | ||
| F1.1.IM.mix v2.0 | Freshwater ecoregions (Abell et al., 2008) were initially identified as containing permanent upland streams if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selected ecoregions were then clipped to exclude cold or dry climates (mean temperature of coldest quarter >0°C, mean annual precipitation >300mm) based on data from Karger et al. (2017). Major occurrences were mapped by intersecting the selected ecoregions with the distribution of 1st-3rd order streams taken from the RiverATLAS (v1.0) database at 30 arc seconds spatial resolution (Linke et al., 2019). The remaining areas within selected ecoregions (excluding freeze/thaw areas) were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.1.web.mix v2.0 | Freshwater ecoregions (Abell et al. 2008) were initially identified as containing permanent upland streams if: i) their descriptions mentioned features consistent with those identifed in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selected ecoregions were then clipped to exlude cold or dry climates (mean temperature of coldest quarter >0°C, mean annual preipitation >300mm) based on data from Karger et al. (2017). Major occurrences were mapped by intersecting the selected ecoregions with the distribution of 1st-3rd order streams taken from the RiverATLAS (v1.0) database at 30 arc seconds spatial resolution (Linke et al. 2019). The remaining areas within selected ecoregions (excluding freeze/thaw areas) were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.2 Permanent lowland rivers | ||
| F1.2.IM.grid v1.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within these areas, major occurrences were mapped using stream orders 4-9 taken from the RiverATLAS (v1.0) database (Linke et al. 2019) combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016). The remaining area of selected ecoregions was designated as minor occurrences. Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F1.2.web.map v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within these areas, major occurrences were mapped using stream orders 4-9 taken from the RiverATLAS (v1.0) database (Linke et al. 2019) combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016). The remaining area of selected ecoregions was designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.3 Freeze-thaw rivers and streams | ||
| F1.3.IM.grid v1.0 | The distribution of freeze-thaw rivers and streams was mapped from the Global River Classification database (Ouellet Dallaire et al., 2018), including all reaches with minimum temperature below 0°C. Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F1.3.web.map v1.0 | The distribution of freeze-thaw rivers and streams was mapped from the Global River Classification database (Ouellet Dallaire et al. 2018), including all reaches with with minimum temperature below 0°C. | JR Ferrer-Paris, DA Keith |
| F1.4 Seasonal upland streams | ||
| F1.4.IM.grid v2.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within selected ecoregions, major occurrences were mapped using 1st-4th order streams (3km buffer) taken from the RiverATLAS (v1.0) database (Linke et al., 2019). The remaining areas of selected ecoregions were mapped as minor occurrences. Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F1.4.IM.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing major or minor occurrences of rivers and streams were identified by consulting available ecoregion descriptions (http://www.feow.org/), global and regional reviews , maps of relevant ecosystems, and expertise of authors. | DA Keith, JR Ferrer-Paris |
| F1.4.web.map v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within the selected ecoregions, major occurrences were mapped using 1st-4th order streams (3km buffer) taken from the RiverATLAS (v1.0) database (Linke et al. 2019). The remaining areas of selected ecoregions were mapped as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.5 Seasonal lowland rivers | ||
| F1.5.IM.grid v2.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within these areas, major occurrences were mapped using stream orders 4-9 taken from the RiverATLAS (v1.0) database (Linke et al. 2019) combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016). The remaining area of selected ecoregions was designated as minor occurrences. Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F1.5.IM.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing major or minor occurrences of rivers and streams were identified by consulting available ecoregion descriptions (http://www.feow.org/), global and regional reviews , maps of relevant ecosystems, and expertise of authors. | DA Keith, JR Ferrer-Paris |
| F1.5.web.map v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within these areas, major occurrences were mapped using stream orders 4-9 taken from the RiverATLAS (v1.0) database (Linke et al.2019) combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016). The remaining area of selected ecoregions was designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.6 Episodic arid rivers | ||
| F1.6.IM.mix v1.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within the selected ecoregions, remote sensing estimates of ephemeral surface water (classes 4, 5 and 8 from Pekel et al., 2016) were used to identify major occurrences at 0.5 minute spatial resolution. The remaining areas of selected ecoregions were mapped as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.6.IM.mix v2.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within the selected ecoregions, remote sensing estimates of seasonal/ephemeral surface water (classes 4, 5 and 8 from Pekel et al. 2016) were used to identify major occurrences. The remaining areas of selected ecoregions were mapped as minor occurrences. We used MERIT Hydro river channels as proxies of the location of rivers and streams within these areas (Yamazaki et al. 2019). | JR Ferrer-Paris, DA Keith |
| F1.6.web.mix v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within the selected ecoregions, remote sensing estimates of ephemeral surface water (classes 4, 5 and 8 from Pekel et al. 2016) were used to identify major occurrences. The remaining areas of selected ecoregions were mapped as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F1.7 Large lowland rivers | ||
| F1.7.IM.orig v2.0 | The distribution of large lowland rivers was taken from the Global River Classification database (Ouellet Dallaire et al., 2018). Reaches with flow > 10,000 m3/s were mapped with a 20 km buffer as major occurrences, clipped to exclude those with seasonal freezing temperatures (mean temperature of coldest quarter <0°C). | JR Ferrer-Paris, DA Keith |
| F1.7.web.orig v2.0 | The distribution of large lowland rivers was taken from the Global River Classification database (Ouellet Dallaire et al. 2018). Reaches with flow > 10,000 m3/s were mapped with a 20 km buffer as major occurrences, clipped to exclude those with seasonal freezing temperatures (mean temperature of coldest quarter <0°C). | JR Ferrer-Paris, DA Keith |
| F2.1 Large permanent freshwater lakes | ||
| F2.1.IM.alt v4.0 | Locations of large lakes (>100km2) were taken from the HydroLAKES database (Messager et al., 2016) and combined with global estimates of permanent surface water surfaces (classes 1, 2 and 7 from Pekel et al., 2016). Freeze/thaw lakes in boreal and polar climates (temperature of coldest quarter > -10°C) were excluded (Beck et al., 2018) (see F2.3). Occurrences were aggregated to 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.1.web.alt v4.0 | Locations of large lakes (>100km2)were taken from the HydroLAKES database (Messager et al. 2016) and combined with global estimates of permanent surface water surfaces (classes 1, 2 and 7 from Pekel et al. 2016). Freeze/thaw lakes in boreal and polar climates (temperature of coldest quarter > -10°C) were excluded (Beck et al. 2018) (see F2.3). Occurrences were aggregated to 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.10 Subglacial lakes | ||
| F2.10.IM.mix v1.0 | Major occurrences of subglacial lakes were mapped as 0.5 degree cells containing the point records of Wright and Siegert (2012), Bowling et al., (2019), Thór Marteinsson et al. (2013) and Livingstone et al. (2016). Unmapped lakes are likely to occur within areas with permanent snow and ice cover and were mapped as minor occurrences based on permanent snow and ice from Dinerstein et al. (2017) and Tuanmu et al. (2014). | JR Ferrer-Paris, DA Keith |
| F2.10.web.mix v2.0 | Major occurrences of subglacial lakes were mapped as point records from Wright and Siegert (2012), Bowler et al. (2019), Marteinsson et al. (2013) and Livingstone et al. (2016). Unmapped lakes are likely to occur within areas with permanent snow and ice cover and were mapped as minor occurrences based on permanent snow and ice from Dinerstein et al. (2017) and Tuanmu et al. (2014). | JR Ferrer-Paris, DA Keith |
| F2.2 Small permanent freshwater lakes | ||
| F2.2.IM.mix v2.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al., 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.2.web.mix v2.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al. 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.3 Seasonal freshwater lakes | ||
| F2.3.IM.mix v2.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al., 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.3.IM.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing major or minor occurrences of rivers and streams were identified by consulting available ecoregion descriptions (http://www.feow.org/), global and regional reviews , maps of relevant ecosystems, and expertise of authors. | DA Keith, JR Ferrer-Paris |
| F2.3.web.mix v2.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al. 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.4 Freeze-thaw freshwater lakes | ||
| F2.4.IM.mix v1.0 | Location of small and large natural lakes was taken from the HydroLAKES database (types 1 an 3 in Messager et al., 2016). We included all lakes with minimum temperature below 0°C (Linke et al., 2019). Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.4.IM.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing major or minor occurrences of rivers and streams were identified by consulting available ecoregion descriptions (http://www.feow.org/), global and regional reviews , maps of relevant ecosystems, and expertise of authors. | DA Keith, JR Ferrer-Paris |
| F2.4.web.mix v1.0 | Location of small and large natural lakes was taken from the HydroLAKES database (types 1 an 3 in Messager et al. 2016). We included all lakes with minimum temperature below 0°C (Linke et al. 2019). Occurrences were aggregated to ten minute spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.5 Ephemeral freshwater lakes | ||
| F2.5.IM.mix v1.0 | Location of natural ephemeral freshwater lakes was taken from global lake databases (Lehner and Döll, 2004; types 1 an 3 from Messager et al., 2016), excluding those from endorheic basins cf. F2.7 (Linke et al., 2019), and intersected with estimates of ephemeral surface water (classes 9 and 10 from Pekel et al., 2016). Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.5.web.mix v1.0 | Location of natural ephemeral freshwater lakes was taken from global lake databases (Lehner and Döll 2004; types 1 an 3 from Messager et al. 2016), excluding those from endorheic basins cf. F2.7 (Linke et al. 2019), and intersected with estimates of ephemeral surface water (classes 9 and 10 from Pekel et al. 2016). Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.6 Permanent salt and soda lakes | ||
| F2.6.IM.mix v4.0 | Major occurrences were compiled from a list of known salt lakes in Wurtsbaugh et al., (2017) and augmented by authors, then matched with names in the HydroLAKES database to identify natural lakes (types 1 and 3 of Messager et al., 2016). Minor occurrences were mapped within arid and semi-arid parts of selected freshwater ecoregions (Abell et al., 2008) by clipping ecoregions to exclude areas with mean annual rainfall >250mm (Harris et al., 2014a). Freshwater ecoregions (Abell et al., 2008) were selected if they contained occurrences of permanent salt or soda lakes if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.6.IM.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing major or minor occurrences of rivers and streams were identified by consulting available ecoregion descriptions (http://www.feow.org/), global and regional reviews , maps of relevant ecosystems, and expertise of authors. | DA Keith, JR Ferrer-Paris |
| F2.6.web.mix v4.0 | Major occurrences were compliled from a list of known salt lakes in Wurtsbaugh et al. (2017) and augmented by authors, then matched with names in the HydroLAKES database to identify natural lakes (types 1 and 3 of Messager et al. 2016). Minor occurrences were mapped within arid and semi-arid parts of selected freshwater ecoregions (Abell et al. 2008) by clipping ecoregions to exclude areas with mean annual rainfall >250mm (Harris et al. 2014a). Freshwater ecoregions (Abell et al. 2008) were selected if they contained occurrences of permanent salt or soda lakes if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.7 Ephemeral salt lakes | ||
| F2.7.IM.mix v1.0 | Location of ephemeral lakes was taken from global lake databases (Lehner and Döll, 2004; types 1 and 3 from Messager et al., 2016), intersected with estimates of ephemeral surface water (classes 9 and 10 from Pekel et al., 2016) and the distribution of arid and semi-arid, endorheic basins (Linke et al., 2019). Occurrences were aggregated to 10 minutes spatial resolution. Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.7.web.mix v1.0 | Location of ephemeral lakes was taken from global lake databases (Lehner and Döll 2004; types 1 and 3 from Messager et al. 2016), intersected with estimates of ephemeral surface water (classes 9 and 10 from Pekel et al. 2016) and the distribution of arid and semi-arid, endorheic basins (Linke et al. 2019). Occurrences were aggregated to 10 minutes spatial resolution. Occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.8 Artesian springs and oases | ||
| F2.8.IM.mix v1.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al., 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.8.web.mix v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al. 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.9 Geothermal pools and wetlands | ||
| F2.9.IM.mix v1.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al., 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al., 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F2.9.web.mix v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within those areas, locations of small lakes (<100km2), excluding artificial lakes (inclusion on types 1 and 3 only), were taken from the HydroLAKES database (Messager et al. 2016) and combined with global estimates of surface water phenology (classes 1, 2 and 7 from Pekel et al. 2016), occurrences were aggregated to 10 minutes spatial resolution. | JR Ferrer-Paris, DA Keith |
| F3.1 Large reservoirs | ||
| F3.1.WM.nwx v1.0 | Polygons of reservoirs were obtained from water bodies in classes 2 and 3 in the HydroLakes database, except for those larger than 100 km2, as these included a number of semi-regulated natural lakes (Messager et al. 2016). Additional point locations were taken from the Global Georeferenced Database of Dams (Mulligan et al. 2020), adding a spatial buffer of 15 arc-minutes to represent uncertainty in their exact location and extent. Major and minor occurrences were not distinguished . | JR Ferrer-Paris |
| F3.1.web.orig v1.0 | Point locations of large reservoirs were obtained from water bodies tagged as 'reservoirs' in "reservoirs" in vector layers GLWD1 and GLWD2 of Lehner & Döll (2004). These were mapped with a spatial buffer of 15 minutes, enabling reservoirs to be represented in 0.5 degree grid cells. | DA Keith, JR Ferrer-Paris |
| F3.2 Constructed lacustrine wetlands | ||
| F3.2.WM.nwx v1.0 | Point locations of over 1 million water bodies (area: 0.1 - 1 km2) were obtained from the HydroLakes database (Messager et al. 2016). We mapped a spatial buffer of 10 km around these locations to generalise their exact location and extent. We estimated the intensity of agricultural use using the mapped area fractions of pasture (PAF) and cropland (CAF) from Ramankutty et al. (2008). We classified the intersection of the lakes-buffer with PAF+CAF>0.5 as major occurrence and the intersection of the lakes-buffer with 0.05<(PAF+CAF)<0.5 as minor occurrences. The resulting map should show the main concentrations of constructed lacustrine wetlands but will underestimate occurrences in non-agricultural areas. | JR Ferrer-Paris, DA Keith |
| F3.2.web.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing major or minor occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identifed in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selections were check by expert reviewers. Occurrences were mapped at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| F3.3 Rice paddies | ||
| F3.3.IM.alt v2.0 | The distribution of rice paddies was estimated from the percentage of rice cover at a 5 arc minute resolution based on Monfreda et al. (2008). Cells with > 10% rice cover were designated as major occurrences, and those with 1-10% rice cover were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F3.3.web.alt v2.0 | The distribution of rice paddies was estimated from the percentage of rice cover at a 5 arc minute resolution based on Monfreda et al. (2008). Cells with > 10% rice cover were designated as major occurrences, and those with 1-10% rice cover were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| F3.4 Freshwater aquafarms | ||
| F3.4.IM.orig v1.0 | Direct data on the distribution of freshwater aquafarms is currently unavailable. To approximate the global distribution, freshwater ecoregions (Abell et al., 2008) were identified as containing major or minor occurrences of freshwater aquafarms if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selections were checked by expert reviewers. Occurrences were mapped at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| F3.4.web.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing major or minor occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identifed in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selections were check by expert reviewers. Occurrences were mapped at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| F3.5 Canals, ditches and drains | ||
| F3.5.WM.nwx v1.0 | We used a global map of irrigation for the year 2015 (Nagaraj et al. 2021) and a global land use/land cover (LULC) map for the year 2020 (Karra et al. 2021). We resampled and aggregated both maps to a 30 arc-second (ca. 1km) resolution. We used the combination of "low-concentration irrigation (1% to 20% irrigated)" class and >1% of built area as minor occurrences, and "high-concentration irrigation (>20% irrigated)" and >1% of built area as major occurrences. | JR Ferrer-Paris, DA Keith |
| F3.5.WM.nwx v2.0 | We used a global map of irrigation areas for the year 2005 () and a global land use/land cover (LULC) map for the year 2020 (Karra et al. 2021). We resampled and aggregated both maps to a 30 arc-second (ca. 1km) resolution. We used a global map of irrigation for the year 2005 (Siebert et al. 2013) to identify irrigation canals and a global land use/land cover (LULC) map for the year 2020 (Karra et al. 2021) to represent drains and canals in built areas. We found fewer mapping artefacts in the irrigation map compiled by Siebert et al. (2013) than a more recent one prepared by Nagaraj et al. (2021). We resampled and aggregated both maps to a 30 arc-second (ca. 1km) resolution. We mapped major occurrences of canals, ditches and drains where the high percentage of area equipped for irrigation was >20% (Siebert et al. 2013) or the proportion of built area was >5% (Karra et al. 2021). We mapped minor occurrences where irrigation-equipped area was 10-20%) or where there was low cover of built area (1-5%). | JR Ferrer-Paris, DA Keith |
| F3.5.web.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) were identified as containing major or minor occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identifed in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. The selections were check by expert reviewers. Occurrences were mapped at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| FM1.1 Deepwater coastal inlets | ||
| FM1.1.IM.alt v3.0 | Known locations of fjords were selected from a global geographical gazetteer (GeoNames, 2020) and the composite gazetteer of Antarctica (SCAR, 1992-2020). We further selected related coastal areas from a global coastal typology (Type IV in Dürr et al., 2011) and the adjacent marine shelves to 2000 metre depth (Becker et al., 2009). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (cells with at least one known occurrence) and minor occurrences (cells with > 5% occurrence of coastal/marine shelf areas). Minor occurrences were clipped to a 50km buffer along the coast to remove inland and oceanic areas. | JR Ferrer-Paris, DA Keith |
| FM1.1.web.alt v3.0 | Known locations of fjords where selected from a global geographical gazetteer (GeoNames 2020) and the composite gazetteer of Antarctica (SCAR 1992-2020). We further selected related coastal areas from a global coastal typology (Type IV in Dürr et al. 2011) and the adjacent marine shelves to 2000 meter depth (Becker et al 2009). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (cells with at least one known occurrence) and minor occurrences (cells with > 5% occurrence of coastal/marine shelf areas). Minor occurrences were clipped to a 50km buffer along the coast to remove inland and oceanic areas. | JR Ferrer-Paris, DA Keith |
| FM1.1.web.map v1.0 | Known locations of fjords where selected from a global geographical gazetteer (GeoNames 2020) and the composite gazetteer of Antarctica (SCAR 1992-2020). We further selected related coastal areas from a global coastal typology (Type IV in Dürr et al. 2011) and the adjacent marine shelves to 2000 meter depth (Becker et al 2009). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (cells with at least one known occurrence) and minor occurrences (cells with > 5% occurrence of coastal/marine shelf areas). Minor occurrences were clipped to a 10km buffer along the coast to remove inland and oceanic areas. | JR Ferrer-Paris, DA Keith |
| FM1.2 Permanently open riverine estuaries and bays | ||
| FM1.2.IM.orig v2.0 | Marine ecoregions (Spalding et al., 2008) containing major or minor occurrences of each Ecosystem Functional Group were identified by consulting global and regional reviews, maps of relevant ecosystems, imagery available in Google Earth and expertise of authors. Occurrences were converted to 30 arc second spatial resolution and clipped to a 50 km buffer along the coastline to exclude inland and offshore areas of the ecoregions. | JR Ferrer-Paris, DA Keith |
| FM1.2.web.map v1.0 | Marine ecoregions (Spalding et al. 2008) containing major or minor occurrences of each ecosystem functional group were identified by consulting global and regional reviews, maps of relevant ecosystems, imagery available in Google Earth and expertise of authors. Occurrences were converted to 30 arc second spatial resolution and clipped to a 20 km buffer along the coastline to exlcude inland and offshore areas of the ecoregions. | JR Ferrer-Paris, DA Keith |
| FM1.3 Intermittently closed and open lakes and lagoons | ||
| FM1.3.IM.orig v2.0 | Marine ecoregions (Spalding et al., 2008) containing major or minor occurrences of each Ecosystem Functional Group were identified by consulting global and regional reviews, maps of relevant ecosystems, imagery available in Google Earth and expertise of authors. Occurrences were converted to 30 arc second spatial resolution and clipped to a 50 km buffer along the coastline to exclude inland and offshore areas of the ecoregions. | JR Ferrer-Paris, DA Keith |
| FM1.3.web.orig v2.0 | Marine ecoregions (Spalding et al. 2008) containing major or minor occurrences of each ecosystem functional group were identified by consulting global and regional reviews, maps of relevant ecosystems, imagery available in Google Earth and expertise of authors. Occurrences were converted to 30 arc second spatial resolution and clipped to a 20 km buffer along the coastline to exlcude inland and offshore areas of the ecoregions. | JR Ferrer-Paris, DA Keith |
| M1.1 Seagrass meadows | ||
| M1.1.IM.orig v1.0 | Indicative maps of Seagrass meadows were obtained from UNEP-WCMC & Short (2017) based on Green & Short (2003). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.1.web.orig v1.0 | Indicative maps of Seagrass meadows were obtained from UNEP-WCMC & Short (2017) based on Green & Short (2003). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.10 Rhodolith/Maërl beds | ||
| M1.10.WM.nwx v1.0 | Distribution of rhodolith beds was taken from spatial predictions of suitable habitats for rhodolith forming species at global scales, including both tropical - warm temperate and polar - cold temperate affiliated rhodolith species (Fragkopoulou et al. 2021). Maps have a spatial resolution of 5 arc-minute. | JR Ferrer-Paris, DA Keith |
| M1.2 Kelp forests | ||
| M1.2.IM.orig v2.1 | Ecoregions with major and minor occurrences of Kelp forests were identified by overlaying a global map of kelp systems (Wernberg and Filbee-Dexter, 2019) on marine ecoregions (Spalding et al., 2008), and then clipping to bathymetry with <80m depth (Becker et al., 2009). Clipped ecoregions were assigned to major and minor occurrences based on information in Wernberg and Filbee-Dexter (2019) and author expertise, and proofed by specialist reviewers. Occurrences were converted to 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| M1.2.web.orig v2.2 | Ecoregions with major and minor occurrences of Kelp forests were identified by overlaying a global map of kelp systems (Wernberg and Filbee-Dexter 2019) on marine ecoregions (Spalding et al. 2008), and then clipping to bathymetry with <80m depth (Becker et al. 2009). Clipped ecoregions were assigned to major and minor occurrences based on information in Wernberg and Filbee-Dexter (2019) and author expertise, and proofed by specialist reviewers. Occurrences were converted to 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| M1.3 Photic coral reefs | ||
| M1.3.IM.orig v1.0 | Indicative maps of Photic coral reefs were obtained from Institute for Marine Remote Sensing et al. (2011). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.3.web.orig v1.0 | Indicative maps of Photic coral reefs were obtained from Institute for Marine Remote Sensing et al. (2011). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.4 Shellfish beds and reefs | ||
| M1.4.IM.orig v1.0 | Major and minor occurrences of shellfish beds and reefs were identified by overlaying a global map of oyster reefs (Beck et al., 2011) on marine ecoregions (Spalding et al., 2008), and then clipping to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.4.web.orig v1.0 | Major and minor occurrences of shellfish beds and reefs were identified by overlaying a global map of oyster reefs (Beck et al. 2011) on marine ecoregions (Spalding et al. 2008), and then clipping to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.5 Photo-limited marine animal forests | ||
| M1.5.IM.orig v1.0 | These are Ecosystem Functional Groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.5.web.orig v1.0 | These are ecosystem functional groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.6 Subtidal rocky reefs | ||
| M1.6.IM.orig v1.0 | These are Ecosystem Functional Groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.6.web.orig v1.0 | These are ecosystem functional groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.7 Subtidal sand beds | ||
| M1.7.IM.orig v1.0 | These are Ecosystem Functional Groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.7.web.orig v1.0 | These are ecosystem functional groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.8 Subtidal mud plains | ||
| M1.8.IM.orig v1.0 | These are Ecosystem Functional Groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.8.web.orig v1.0 | These are ecosystem functional groups that are widespread through the global extent of the marine shelf biome. Reliable data on their precise distribution are limited. To represent regional uncertainty, their indicative distributions were mapped in as minor occurrences through the full extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.9 Upwelling zones | ||
| M1.9.IM.orig v1.0 | Marine ecoregions (Spalding et al., 2008) with major and minor occurrences of Upwelling zones were identified by consulting global and regional reviews (cited in descriptive profile), maps of relevant ecosystems and expertise of authors, proofed by specialist reviewers. The identified ecoregions were then clipped to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M1.9.web.orig v1.0 | Marine ecoregions (Spalding et al. 2008) with major and minor occurrences of Upwelling zones were identified by consulting global and regional reviews (cited in descriptive profile), maps of relevant ecosystems and expertise of authors, proofed by specialist reviewers. The identified ecoregions were then clipped to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.1 Epipelagic ocean waters | ||
| M2.1.IM.orig v2.1 | Indicative distributions of these marine pelagic Ecosystem Functional Groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| M2.1.web.orig v2.1 | Indicative distributions of these marine pelagic ecosystem functional groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| M2.2 Mesopelagic ocean water | ||
| M2.2.IM.orig v2.0 | Indicative distributions of these marine pelagic Ecosystem Functional Groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.2.web.orig v2.0 | Indicative distributions of these marine pelagic ecosystem functional groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.3 Bathypelagic ocean waters | ||
| M2.3.IM.orig v1.0 | Indicative distributions of these marine pelagic Ecosystem Functional Groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.3.web.orig v1.0 | Indicative distributions of these marine pelagic ecosystem functional groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.4 Abyssopelagic ocean waters | ||
| M2.4.IM.orig v1.0 | Indicative distributions of these marine pelagic Ecosystem Functional Groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.4.web.orig v1.0 | Indicative distributions of these marine pelagic ecosystem functional groups were derived from bathymetric spatial data obtained from Becker et al. (2009) using depth range thresholds cited in respective descriptive profiles for each functional group. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.5 Sea ice | ||
| M2.5.IM.orig v1.0 | Indicative distributions of sea ice were obtained from Fetterer et al. (2017). To approximate the maximum annual global extent, we used the monthly extent for March 2019 for the northern hemisphere, and the monthly extent for September 2018 for the southern hemisphere. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M2.5.web.orig v1.0 | Indicative distributions of sea ice were obtained from Fetterer et al. (2017). To approximate the maximum annual global extent, we used the monthly extent for March 2019 for the northern hemisphere, and the monthly extent for September 2018 for the southern hemisphere. Occurrences were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.1 Continental and island slopes | ||
| M3.1.IM.orig v1.0 | Major occurrences of continental and island slopes was based on the ‘slope’ geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.1.web.orig v1.0 | Major occurrences of continental and island slopes was based on the ‘slope’ geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.2 Submarine canyons | ||
| M3.2.IM.orig v1.0 | Major occurrences of submarine canyons was based on the ‘canyons’ geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.2.web.orig v1.0 | Major occurrences of submarine canyons was based on the ‘canyons’ geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.3 Abyssal plains | ||
| M3.3.IM.orig v1.0 | Major occurrences of Abyssal plains was based on the ‘plains’ and ‘hills’ classes within the abyssal geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.3.web.orig v1.0 | Major occurrences of Abyssal plains was based on the ‘plains’ and ‘hills’ classes within the abyssal geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.4 Seamounts, ridges and plateaus | ||
| M3.4.IM.orig v2.0 | Major occurrences of seamounts, ridges and plateaus was based on the ‘mountains’ classes within the abyssal geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.4.web.orig v2.0 | Major occurrences of seamounts, ridges and plateaus was based on the ‘mountains’ classes within the abyssal geomorphic unit of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.5 Deepwater biogenic beds | ||
| M3.5.IM.orig v1.0 | The distribution of deepwater biogenic beds was based on the ‘mountains’ and ‘hills’ classes within the abyssal geomorphic unit of Harris et al. (2014b). These were mapped in yellow as minor occurrences to acknowledge considerable uncertainties in the distribution of biogenic beds within these geomorphic units. Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.5.web.orig v1.0 | The distribution of deepwater biogenic beds was based on the ‘mountains’ and ‘hills’ classes within the abyssal geomorphic unit of of Harris et al. (2014b). These were mapped in yellow as minor occurrences to acknowledges considerable uncertainties in the distribution of biogenic beds within these geomorphic units. Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.6 Hadal trenches and troughs | ||
| M3.6.IM.orig v1.0 | Major occurrences of Hadal trenches and troughs were based on the ‘hadal’ and ‘trenches’ geomorphic units of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.6.web.orig v1.0 | Major occurrences of Hadal trenches and troughs was based on the ‘hadal’ and ‘trenches’ geomorphic units of Harris et al. (2014b). Occurrences were converted to 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| M3.7 Chemosynthetic-based-ecosystems (CBE) | ||
| M3.7.IM.orig v1.0 | Major occurrences of Chemosynthetic-based ecosystems were based on the distribution of hydrothermal vents on spreading plate boundaries mapped in ‘Plate lines and polygons’ data by USGS/ESRI (undated). Occurrences were converted to 30 arc second spatial resolution. The distribution of cold seeps is poorly known and was not mapped. | DA Keith, JR Ferrer-Paris |
| M3.7.web.orig v1.0 | Major occurrences of Chemosynthetic-based ecosystems was based on the distribution of hydrothermal vents on spreading plate boundaries mapped in ‘Plate lines and polygons’ data by USGS/ESRI. Occurrences were converted to 30 arc second spatial resolution. The distribution of cold seeps is poorly known and was not mapped. | DA Keith, JR Ferrer-Paris |
| M4.1 Submerged artificial structures | ||
| M4.1.IM.orig v1.0 | Marine ecoregions that include occurrences of submerged artificial structures were identified by overlaying a mapped distribution of shipwrecks (Monfils, 2004) on marine ecoregions (Spalding et al., 2008). Occurrences were converted to 30 arc second spatial resolution. In many cases these ecoregions encompassed other submerged structures such as energy infrastructure. To represent uncertainty, indicative distributions were mapped as minor occurrences. | DA Keith, JR Ferrer-Paris |
| M4.1.web.orig v1.0 | Marine ecoregions that include occurrences of submerged artificial structures were identified by overlaying a mapped distribution of shipwrecks (Monfils 2004) on marine ecoregions (Spalding et al. 2008). Occurrences were converted to 30arc second spatial resolution. In many cases these ecoregions encompassed other submerged structures such as energy infrastructure. To represent uncertainty, indicative distributions were mapped as minor occurrences. | DA Keith, JR Ferrer-Paris |
| M4.2 Marine aquafarms | ||
| M4.2.IM.orig v1.0 | Marine ecoregions (Spalding et al., 2008) containing marine aquafarms were identified by consulting global and regional reviews, suitability maps (Gentry et al., 2017) and expertise of authors, proofed by specialist reviewers. These were clipped to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b) and converted to 30 arc second spatial resolution. Occurrences were aggregated to half degree spatial resolution. | DA Keith, JR Ferrer-Paris |
| M4.2.web.orig v1.0 | Marine ecoregions (Spalding et al. 2008) containing marine aquafarms were identified by consulting global and regional reviews, suitability maps (Gentry et al. 2017) and expertise of authors, proofed by specialist reviewers. These were clipped to the extent of the marine ‘shelf’ base layer as mapped by Harris et al. (2014b) and converted to 30 arc second spatial resolution. Occurrences were aggregated to half degree spatial resolution. | DA Keith, JR Ferrer-Paris |
| MFT1.1 Coastal river deltas | ||
| MFT1.1.WM.nwx v1.0 | The extent of major coastal deltas was taken directly from Nienhuis et al. (2020). The data are based on polygons that encompass the lowest reaches of deltaic floodplains and a marine buffer approximating the extent of subtidal deltaic sediments. We checked the data for completeness against point locations shown in Fig. 1 of Goodbred & Saito (2012) and maps of Tessler et al. (2015) and found them to be inclusive of major occurrences. The latter included fewer deltas and olygons that extended some distance up freshwater floodplains into the Freshwater-Terrestrial (FT) transition biome. | JR Ferrer-Paris |
| MFT1.1.web.orig v1.0 | Point locations for major coastal deltas were mapped directly from data used for Fig. 1 in Goodbred & Saito (2012). | DA Keith, JR Ferrer-Paris |
| MFT1.2 Intertidal forests and shrublands | ||
| MFT1.2.IM.orig v0.2 | The indicative map of Intertidal forests and shrublands (MFT1.2) was developed by resampling the known global distribution of mangrove forests for the year 2000 mapped by Giri et al. (2011). We used a buffer of 1km around the distribution data and a 30 arc second grid, thus large aggregations (> 1km2) are depicted as major occurrences, and the buffer areas with small occurrences are shown as minor occurrences. | JR Ferrer-Paris, DA Keith |
| MFT1.2.WM.nwx v1.0 | The indicative map for Intertidal forests and shrublands was was developed by resampling the known global distribution of mangrove forests for the year 2016 mapped by Global Mangrove Watch (Bunting et al. 2018). We used a buffer of 1km around the distribution data and a 30 arc second grid, thus large aggregations (> 1km2) are depicted as major occurrences, and the buffer areas with small occurrences are shown as minor occurrences. | JR Ferrer-Paris |
| MFT1.3 Coastal saltmarshes and reedbeds | ||
| MFT1.3.IM.orig v1.0 | The indicative map for Coastal saltmarshes was based on mapping by McOwen et al. (2017) summarised within a template of 1-degree grid cells. Cells with >5% cover of marsh vegetation were reclassified as major occurrences, and those with non-zero cover up to 5% were reclassified as minor occurrences. | DA Keith, JR Ferrer-Paris |
| MFT1.3.IM.orig v2.0 | The indicative map for for coastal saltmarshes and reedbeds (MFT1.3) was developed by resampling known distribution records summarised by McOwen et al. (2017). We used a buffer of 1km around the distribution data and a 30 arc second grid, thus large aggregations (> 1km2) are depicted as major occurrences, and the buffer areas with small occurrences are shown as minor occurrences. | JR Ferrer-Paris, DA Keith |
| MFT1.3.web.orig v2.0 | The indicative map for for coastal saltmarshes and reedbeds (MFT1.3) was developed by resampling known distribution records summarised by McOwen et al. (2017). We used a buffer of 1km around the distribution data and a 30 arc second grid, thus large aggregations (> 1km2) are depicted as major occurrences, and the buffer areas with small occurrences are shown as minor occurrences. The original data is available at [UNEP-WCMC](http://data.unep-wcmc.org/datasets/43). | JR Ferrer-Paris, DA Keith |
| MT1.1 Rocky Shorelines | ||
| MT1.1.IM.grid v2.0 | Marine ecoregions (Spalding et al., 2008) containing rocky shorelines and boulder and cobble shorelines, respectively, were identified by consulting regional substrate maps, imagery available in Google Earth (to exclude ecoregions with extensive sandy or muddy shores) and expertise of authors, proofed by specialist reviewers. Occurrences were aggregated to 1 degree spatial resolution. | JR Ferrer-Paris, DA Keith |
| MT1.1.web.map v1.0 | Marine ecoregions (Spalding et al. 2008) containing rocky shorelines and boulder and cobble shorelines, respectively, were identified by consulting regional substrate maps, imagery available in Google Earth (to exclude ecoregions with extensive sandy or muddy shores) and expertise of authors, proofed by specialist reviewers. | JR Ferrer-Paris, DA Keith |
| MT1.2 Muddy Shorelines | ||
| MT1.2.IM.orig v1.0 | Tidal flats were mapped directly from remote sensing time series and aggregated to 1 degree spatial resolution by Murray et al. (2019). Major occurrences were mapped in 1-degree cells with >200km2 mudflat extent, and minor occurrences were mapped in cells with 5-200km2 mudflat extent. | NJ Murray, DA Keith, JR Ferrer-Paris |
| MT1.2.web.map v1.0 | The indicative map of muddy shorelines was developed by resampling the 2016 area of tidal flats mapped by Murray et al. (2019) at two resolutions. A broad grid cell (5 arc minute resolution) is used to show the overall range of tidal flats, and large aggregations are depicted as major occurrences (30 arc second grids or approx. 1km2). | NJ Murray, JR Ferrer-Paris, DA Keith |
| MT1.3 Sandy Shorelines | ||
| MT1.3.IM.grid v1.0 | The indicative map of Sandy shorelines was based on point records of sandy coastlines mapped by Vousdoukas et al. (2020) aggregated to 1 degree spatial resolution. Cells with >50 points were reclassified as major occurrences, and those with 1-50 points were reclassified as minor occurrences. | JR Ferrer-Paris, DA Keith |
| MT1.3.web.map v1.0 | The indicative map of Sandy shorelines was based on point records of sandy coastlines mapped by Vousdoukas et al. (2020) aggregated to 1 degree spatial resolution. Cells with >50 points were reclassified as major occurrences, and those with 1-50 points were reclassified as minor occurrences. These were clipped to the coastline. | JR Ferrer-Paris, DA Keith |
| MT1.4 Boulder and cobble shores | ||
| MT1.4.IM.grid v2.0 | Marine ecoregions (Spalding et al., 2008) containing rocky shorelines and boulder and cobble shorelines, respectively, were identified by consulting regional substrate maps, imagery available in Google Earth (to exclude ecoregions with extensive sandy or muddy shores) and expertise of authors, proofed by specialist reviewers. Occurrences were aggregated to 1 degree spatial resolution. | JR Ferrer-Paris, DA Keith |
| MT1.4.web.map v1.0 | Marine ecoregions (Spalding et al. 2008) containing rocky shorelines and boulder and cobble shorelines, respectively, were identified by consulting regional substrate maps, imagery available in Google Earth (to exclude ecoregions with extensive sandy or muddy shores) and expertise of authors, proofed by specialist reviewers. | JR Ferrer-Paris, DA Keith |
| MT2.1 Coastal shrublands and grasslands | ||
| MT2.1.IM.orig v1.0 | Coastlines were mapped between 60°S and 60°N with a 20 km buffer applied. | DA Keith, JR Ferrer-Paris |
| MT2.1.web.map v1.0 | Coastlines were mapped with a 5 km buffer applied between 60°S and 60°N. | JR Ferrer-Paris, DA Keith |
| MT2.2 Large seabird and pinniped colonies | ||
| MT2.2.IM.orig v1.0 | We used spatial data on Nitrogen (N) and Phosphor (P) deposition from seabird colonies (Otero et al. 2018) as an indicators of the distribution of this functional group. Original point data was in decimal degrees rounded to 6 arc-min resolution, for the maps we aggregated data to square grid cells of 250 km. We used a threshold of >1000 and <100000 kg/yr N to select minor occurrences and a threshold of >100000 kg/yr N for major occurrences. | JR Ferrer-Paris, S Gorta, DA Keith |
| MT3.1 Artificial shorelines | ||
| MT3.1.IM.grid v2.0 | Marine ecoregions (Spalding et al., 2008) containing major and minor occurrences of urbanised shorelines were identified from the map of night lights (Cinzano et al. 2019), imagery available on Google Earth and expertise of authors. Occurrences were aggregated to 1 degree spatial resolution and intersected with the coastline to exclude areas inland and in the open ocean. | JR Ferrer-Paris, DA Keith |
| MT3.1.web.map v1.0 | Marine ecoregions (Spalding et al. 2008) containing major and minor occurrences of urbanized shorelines were identified from the map of night lights (see T7.4), imagery available on Google Earth and expertise of authors. Occurrences were aggregated to 1 degree spatial resolution and intersected with the coastline to exclude areas inland and in the open ocean. | JR Ferrer-Paris, DA Keith |
| S1.1 Aerobic caves | ||
| S1.1.IM.orig v1.0 | Distributions of Aerobic caves were based on mapped area of carbonate rock outcrop (Williams & Ting Fong 2016). This. provides an upper limit on the area of exposed karst terrain, as not all carbonate rocks are karstified. Lava tubes and other rocks that may contain these ecosystem functional groups are not shown on this indicative map, but are less extensive than those in carbonate rock. | DA Keith, JR Ferrer-Paris |
| S1.1.web.orig v1.0 | Distributions of Aerobic caves were based on mapped area of carbonate rock outcrop (Williams & Ting Fong 2016). This. provides an upper limit on the area of exposed karst terrain, as not all carbonate rocks are karstified. Lava tubes and other rocks that may contain these ecosystem functional groups are not shown on this indicative map, but are less extensive than those in carbonate rock. | DA Keith, JR Ferrer-Paris |
| S1.2 Endolithic systems | ||
| S1.2.IM.orig v1.0 | Global distribution throughout the earth's crust. Not mapped. | DA Keith, JR Ferrer-Paris |
| S1.2.web.orig v1.0 | Global distribution throughout the earth's crust. Not mapped. | DA Keith, JR Ferrer-Paris |
| S2.1 Anthropogenic subterranean voids | ||
| S2.1.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| S2.1.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| SF1.1 Underground streams and pools | ||
| SF1.1.IM.orig v1.0 | Distributions of Underground streams and pools were based on mapped area of carbonate rock outcrop (Williams & Ting Fong 2016). This. provides an upper limit on the area of exposed karst terrain, as not all carbonate rocks are karstified. Lava tubes and other rocks that may contain these ecosystem functional groups are not shown on this indicative map, but are less extensive than those in carbonate rock. | DA Keith, JR Ferrer-Paris |
| SF1.1.web.orig v1.0 | Distributions of Underground streams and pools were based on mapped area of carbonate rock outcrop (Williams & Ting Fong 2016). This. provides an upper limit on the area of exposed karst terrain, as not all carbonate rocks are karstified. Lava tubes and other rocks that may contain these ecosystem functional groups are not shown on this indicative map, but are less extensive than those in carbonate rock. | DA Keith, JR Ferrer-Paris |
| SF1.2 Groundwater ecosystems | ||
| SF1.2.IM.orig v1.0 | Indicative global maps of Groundwater aquifers were based on Bundesanstalt für Geowissenschaften und Rohstoffe & UNESCO (2012) with colour ramp showing type of aquifer by recharge rate, only in major groundwater basins (type 11 (minor occurrences) to type 15 (major occurrences)). | DA Keith, JR Ferrer-Paris |
| SF1.2.web.orig v1.0 | Indicative global maps of Groundwater aquifers were based on Bundesanstalt für Geowissenschaften und Rohstoffe & UNESCO (2012) with colour ramp showing type of aquifer by recharge rate, only in major groundwater basins (type 11 (minor occurrences) to type 15 (major occurrences)). | DA Keith, JR Ferrer-Paris |
| SF2.1 Water pipes and subterranean canals | ||
| SF2.1.IM.orig v1.0 | Freshwater ecoregions (Abell et al., 2008) containing urban and industrialised areas with water transfer infrastructure were identified by consulting available ecoregion descriptions (http://www.feow.org/), maps of irrigation and other water infrastructure, and expertise of authors. Due to uncertainty and limited verification and likely limited spatial extent within mapped areas, all inferred occurrences were shown as minor at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| SF2.1.web.orig v1.0 | Freshwater ecoregions (Abell et al. 2008) containing urban and industrialised areas with water transfer infrastructure were identified by consulting available ecoregion descriptions (http://www.feow.org/), maps of irrigation and other water infrastructure, and expertise of authors. Due to uncertainty and limited verification and likely limited spatial extent within mapped areas, all inferred occurrences were shown as minor at 30 arc seconds spatial resolution. | DA Keith, JR Ferrer-Paris |
| SF2.2 Flooded mines and other voids | ||
| SF2.2.IM.orig v1.0 | Point records of flooded mines were compiled from public databases (https://www.unexmin.eu/the-european-inventory-of-flooded-mines-is-now-online/), an internet search for "flooded mines" and locations of deep mines inferred from world mineral resources spatial data (USGS: https://mrdata.usgs.gov/). Terrestrial ecoregions (Dinerstein et al., 2017) with concentrations of these records were selected to represent an indicative global distribution of flooded mines at 30 arc seconds spatial resolution. | DA Keith |
| SF2.2.web.orig v1.0 | Point records of flooded mines were compiled from public databases (https://www.unexmin.eu/the-european-inventory-of-flooded-mines-is-now-online/), an internet search for "flooded mines" and locations of deep mines inferred from world mineral resources spatial data (USGS: https://mrdata.usgs.gov/). Terrestrial ecoregions (Dinerstein et al. 2017) with concentrations of these records were selected to represent an indicative global distribution of flooded mines at 30 arc seconds spatial resolution. | DA Keith |
| SM1.1 Anchialine caves | ||
| SM1.1.IM.grid v3.0 | Indicative distributions of anchialine caves and pools were based on mapped areas of carbonate rock outcrop (Williams & Ting Fong, 2016) and lava flows intersecting the coast, which were aggregated within a template of 1-degree grid cells. | JR Ferrer-Paris, DA Keith |
| SM1.1.web.grid v3.0 | Indicative distributions of Anchialine caves and pools were based on mapped areas of carbonate rock outcrop (Williams & Ting Fong 2016) and lava flows intersecting the coast, which were aggregated within a template of 1-degree grid cells. | JR Ferrer-Paris, DA Keith |
| SM1.2 Anchialine pools | ||
| SM1.2.IM.grid v3.0 | Indicative distributions of anchialine caves and pools were based on mapped areas of carbonate rock outcrop (Williams & Ting Fong, 2016) and lava flows intersecting the coast, which were aggregated within a template of 1-degree grid cells. | JR Ferrer-Paris, DA Keith |
| SM1.2.web.grid v3.0 | Indicative distributions of Anchialine caves and pools were based on mapped areas of carbonate rock outcrop (Williams & Ting Fong 2016) and lava flows intersecting the coast, which were aggregated within a template of 1-degree grid cells. | JR Ferrer-Paris, DA Keith |
| SM1.3 Sea caves | ||
| SM1.3.IM.grid v3.0 | Marine ecoregions (Spalding et al., 2008) containing occurrences of rocky coastline (see MT1.1) were verified by inspection of imagery available in Google Earth to identify an envelope of potential distribution for sea caves. The coastlines within these ecoregions were summarised using a template of 1-degree grid cell intersected with the coast. As caves represent a small portion of such coastlines, all mapped areas were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| SM1.3.web.grid v3.0 | Marine ecoregions (Spalding et al. 2008) containing occurrences of rocky coastline (see MT1.1) were verified by inspection of imagery available in Google Earth to identify an envelope of potential distribution for sea caves. The coastlines within these ecoregions were summarised using a template of 1-degree grid cell intersected with the coast. As caves represent a small portion of such coastlines, all mapped areas were designated as minor occurrences. | JR Ferrer-Paris, DA Keith |
| T1.1 Tropical/Subtropical lowland rainforests | ||
| T1.1.IM.mix v2.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T1.1.web.mix v2.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017; DAWE 2012) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T1.2 Tropical/Subtropical dry forests and thickets | ||
| T1.2.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T1.2.IM.orig v1.0 | Terrestrial ecoregions (Dinerstein et al. 2017) containing major or minor occurrences of each ecosystem functional group were identified by consulting available ecoregion descriptions (https://www.worldwildlife.org/biome-categories/terrestrial-ecoregions), global and regional reviews (cited in respective reference sections of profiles), maps of relevant ecosystems, and experts. | DA Keith, JR Ferrer-Paris |
| T1.2.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T1.3 Tropical/Subtropical montane rainforests | ||
| T1.3.IM.alt v2.1 | The distribution of tropical montane rainforest was approximated from a model of environmental suitability based on climatic variables and cloud cover (Wilson and Jetz, 2016). Occurrences were aggregated to half degree spatial resolution and cells reclassified as major occurrences (>25% of cell area) and minor occurrences (< 25% of cell area). | JR Ferrer-Paris, DA Keith |
| T1.3.WM.nwx v1.0 | The distribution of tropical montane rainforest was approximated from a model of environmental suitability based on climatic variables and cloud cover (Wilson and Jetz, 2016, Karger et al. 2021). Occurrences were aggregated to half degree spatial resolution and cells reclassified as major occurrences (>25% of cell area) and minor occurrences (< 25% of cell area). | JR Ferrer-Paris |
| T1.3.web.alt v2.1 | The distribution of tropical montane rainforest was approximated from a model of environmental suitability based on climatic variables and cloud cover (Wilson and Jetz 2016). Occurrences were aggregated to half degree spatial resolution and cells reclassified as major occurrences (>25% of cell area) and minor occurrences (< 25% of cell area). | JR Ferrer-Paris, DA Keith |
| T1.4 Tropical heath forests | ||
| T1.4.IM.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T1.4.web.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T2.1 Boreal and temperate high montane forests and woodlands | ||
| T2.1.IM.alt v2.1 | The distribution for this functional group was approximated from a model of the spatial distribution of terrestrial habitat types (Jung et al. 2020). Occurrences were aggregated to half degree spatial resolution and cells reclassified as major occurrences (>25%) and minor occurrences (< 25%) . | JR Ferrer-Paris, DA Keith |
| T2.1.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T2.1.IM.orig v1.0 | Terrestrial ecoregions (Dinerstein et al. 2017) containing major or minor occurrences of each ecosystem functional group were identified by consulting available ecoregion descriptions (https://www.worldwildlife.org/biome-categories/terrestrial-ecoregions), global and regional reviews, maps of relevant ecosystems, and experts. | DA Keith, JR Ferrer-Paris |
| T2.1.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T2.2 Deciduous temperate forests | ||
| T2.2.IM.alt v2.0 | Areas of deciduous and mixed forest cover were identified from consensus land-cover maps (Tuanmu et al. 2014), and cropped to areas with temperate climates in the northern hemisphere (Karger et al. 2017), occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (>80% occurrence) and minor occurrences (< 80% occurrence). | JR Ferrer-Paris, DA Keith |
| T2.2.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T2.2.IM.orig v1.0 | Terrestrial ecoregions (Dinerstein et al. 2017) containing major or minor occurrences of each ecosystem functional group were identified by consulting available ecoregion descriptions (https://www.worldwildlife.org/biome-categories/terrestrial-ecoregions), global and regional reviews (cited in respective reference sections of profiles), maps of relevant ecosystems, and experts. | DA Keith, JR Ferrer-Paris |
| T2.2.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T2.3 Oceanic cool temperate rainforests | ||
| T2.3.IM.orig v1.0 | Cool temperate and boreal rainforest regions identified by DellaSala et al. (2011) matched with ecoregion descriptions (Dinerstein et al., 2017) to map major occurrences, proofed by expert appraisal. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T2.3.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T2.4 Warm temperate laurophyll forests | ||
| T2.4.IM.mix v1.0 | Major and minor occurrences were identified using consensus land-cover maps (Tuanmu et al. 2014) cropped to selected terrestrial ecoregions (Dinerstein et al. 2017). | JR Ferrer-Paris, DA Keith |
| T2.4.IM.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith |
| T2.4.web.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith |
| T2.5 Temperate pyric humid forests | ||
| T2.5.IM.alt v2.0 | Remote sensing estimates of canopy height were used as a direct indicator of the distribution of this group of tall forest ecosystems (Armston et al., 2015, Tang et al., 2019). We selected all areas with tree canopies taller than 40m, and clipped to the spatial extent of temperate climate types (Beck et al., 2018). Mapped occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (>20% of cell area) and minor occurrences (< 20% of cell area). | JR Ferrer-Paris, DA Keith |
| T2.5.web.alt v2.0 | Remote sensing estimates of canopy height were used as a direct indicator of the distribution of this group of tall forest ecosystems (Armston et al. 2015, Hao et al. 2019). We selected all areas with tree canopies taller than 40m, and clipped to the spatial extent of temperate climate types (Beck et al. 2018). Mapped occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (>20% of cell area) and minor occurrences (< 20% of cell area). | JR Ferrer-Paris, DA Keith |
| T2.6 Temperate pyric sclerophyll forests and woodlands | ||
| T2.6.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T2.6.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T3.1 Seasonally dry tropical shrublands | ||
| T3.1.IM.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T3.1.web.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T3.2 Seasonally dry temperate heath and shrublands | ||
| T3.2.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T3.2.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T3.3 Cool temperate heathlands | ||
| T3.3.IM.alt v4.0 | Major and minor occurrences were identified using consensus land-cover maps (Tuanmu et al., 2014; Latifovic et al., 2016), then cropped to selected terrestrial ecoregions at 30 arc seconds spatial resolution (Dinerstein et al., 2017; CEC 1997). Ecoregions were selected if they contained areas mentioned or mapped in published regional studies (Loidi et al., 2015; Luebert & Pliscoff, 2017), or if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T3.3.web.alt v4.0 | Major and minor occurrences were identified using consensus land-cover maps (Tuanmu et al. 2014; Latifovic et al. 2016), then cropped to selected terrestrial ecoregions at 30 arc seconds spatial resolution (Dinerstein et al. 2017; CEC 1997). Ecoregions were selected if they contained areas mentioned or mapped in published regional studies (Loidi et al. 2015; Luebert & Pliscoff 2017), or if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T3.4 Young rocky pavements, lava flows and screes | ||
| T3.4.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T3.4.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T4.1 Trophic savannas | ||
| T4.1.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T4.1.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T4.2 Pyric tussock savannas | ||
| T4.2.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.2.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.3 Hummock savannas | ||
| T4.3.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.3.WM.nwx v1.0 | The distribution of Hummock savannas in Australia was compiled from multiple source maps (Keith and Tozer 2017). It was mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T4.3.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.4 Temperate woodlands | ||
| T4.4.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.4.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.5 Temperate subhumid grasslands | ||
| T4.5.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T4.5.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T5.1 Semi-desert steppe | ||
| T5.1.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.1.IM.orig v1.0 | TEOW | DA Keith, JR Ferrer-Paris |
| T5.1.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.2 Succulent or Thorny deserts and semi-deserts | ||
| T5.2.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.2.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.3 Sclerophyll hot deserts and semi-deserts | ||
| T5.3.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.3.IM.orig v1.0 | Terrestrial ecoregions (Dinerstein et al. 2017) containing major or minor occurrences of each ecosystem functional group were identified by consulting available ecoregion descriptions (https://www.worldwildlife.org/biome-categories/terrestrial-ecoregions), global and regional reviews, maps of relevant ecosystems, and experts. | DA Keith, JR Ferrer-Paris |
| T5.3.WM.nwx v1.0 | The distribution of Sclerophyll hot deserts and semi-deserts in Australia was compiled from multiple source maps (Keith and Tozer 2017). It was mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T5.3.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.4 Cool deserts and semi-deserts | ||
| T5.4.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.4.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.5 Hyper-arid deserts | ||
| T5.5.IM.alt v2.0 | We calculated the aridity index (AI, UNEP 1992) using spatio-temporal series of precipitation and potential evapotranspiration (Harris et al. 2014), and selected areas of hyper-arid climate (AI<0.05). | JR Ferrer-Paris, DA Keith |
| T5.5.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T5.5.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T6.1 Ice sheets, glaciers and perennial snowfields | ||
| T6.1.IM.alt v2.0 | Areas of permanent snow were identified from consensus land-cover maps (Tuanmu et al., 2014), glacier inventories (Raup et al., 2007; GLIMS and NSIDC, 2005-2018) and the Antarctic Land Cover map for 2000 (Hui et al., 2017). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified as major occurrences (cells with > 22% snow coverage) and minor occurrences (cells with at least one occurrence). | JR Ferrer-Paris, DA Keith |
| T6.1.web.map v1.0 | Areas of permanent snow where identified from consensus land-cover maps (Tuanmu et al. 2014), glacier inventories (Raup et al 2007; GLIMS and NSIDC 2005-2018) and the Antarctic Land Cover map for 2000 (Hui et al. 2017). A composite map was created at 30 arc seconds spatial resolution in geographic projection. | JR Ferrer-Paris, DA Keith |
| T6.2 Polar/alpine cliffs, screes, outcrops and lava flows | ||
| T6.2.IM.alt v2.0 | Known locations of prominent ice-free rock in glacial and alpine environments were selected from global geographical gazeteers (GeoNames, 2020), glacier inventories (Raup et al 2007; GLIMS and NSIDC, 2005-2018) and the Antarctic Land Cover map for 2000 (Hui et al., 2017). Further areas with mixed occurrence of barren and snow/ice cover were identified from the Circumpolar Arctic Vegetation Map (Raynolds et al., 2019), the USGS EROS LandCover GLCCDB, version 2 (Loveland et al., 2000) and a 1km consensus land-cover map (Tuanmu et al., 2014). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree cells. Cells containing at least one known location were designated major occurrences, while those mapped as mixed barren and snow/ice cover were designated as minor occurrences if snow/ice covered at least 2.5% of the cell area. | JR Ferrer-Paris, DA Keith |
| T6.2.web.alt v2.0 | Known locations of prominent ice-free rock in glacial and alpine environments were selected from global geographical gazeteers (GeoNames 2020), glacier inventories (Raup et al 2007; GLIMS and NSIDC 2005-2018) and the Antarctic Land Cover map for 2000 (Hui et al. 2017). Further areas with mixed occurrence of barren and snow/ice cover were identified from the Circumpolar Arctic Vegetation Map (Raynolds et al. 2019), the USGS EROS LandCover GLCCDB, version 2 (Loveland et al. 2000) and a 1km consensus land-cover map (Tuanmu et al. 2014). A composite map was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree cells. Cells containing at least one known location were designated as major occurrences, while those mapped as mixed barren and snow/ice cover were designated as minor occurrences if snow/ice covered at least 2.5% of the cell area. | JR Ferrer-Paris, DA Keith |
| T6.3 Polar tundra and deserts | ||
| T6.3.IM.alt v2.0 | Areas corresponding to the tundra climatic zone according to the Köppen-Geiger classification system (Beck et al., 2018) were first identified. Additional areas were then selected in high latitudes corresponding with low annual solar radiation (values <1800 in Beckmann et al., 2014). A union of these maps was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified cells as major occurrences (>80% of cell area) and minor occurrences (30-80% of cell area). | JR Ferrer-Paris, DA Keith |
| T6.3.web.alt v2.0 | Areas corresponding to the tundra climatic zone according to the Köppen-Geiger classification system (Beck et al. 2018) were first identified. Additional areas were then selected in high latitudes corresponding with low annual solar radiation (values <1800 in Beckmann et al. 2014). A union of these maps was created at 30 arc seconds spatial resolution in geographic projection, occurrences were then aggregated to half degree spatial resolution and reclassified cells as major occurrences (>80% of cell area) and minor occurrences (30-80% of cell area). | JR Ferrer-Paris, DA Keith |
| T6.3.web.map v1.0 | Areas corresponding to the tundra climatic zone according to the Köppen-Geiger classification system (Beck et al. 2018) were first identified. Additional areas were then selected in high latitudes corresponding with low annual solar radiation (values <1800 in Beckmann et al. 2014). A union of these maps was created at 30 arc seconds spatial resolution in geographic projection. | JR Ferrer-Paris, DA Keith |
| T6.4 Temperate alpine grasslands and shrublands | ||
| T6.4.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T6.4.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| T6.5 Tropical alpine grasslands and herbfields | ||
| T6.5.IM.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al., 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al., 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T6.5.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| T7.1 Annual croplands | ||
| T7.1.IM.alt v2.0 | Major occurrences of croplands were taken from the map of Habitat type 14.1 by Jung et al. (2020) based on the IUCN Habitats Classification Scheme v3.1 (IUCN 2012). We compared this to cropping areas in consensus land-cover maps (Tuanmu et al., 2014) and found that maps of Jung et al. (2020) more closely matched the concept of T7.1. Occurrences were extracted from fractional aggregated 1 km resolution base data (Jung et al. 2020), approximating 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T7.1.web.alt v2.0 | Major occurrences of croplands were taken from the map of Habitat type 14.1 by Jung et al. (2020) based on the IUCN Habitats Classification Scheme v3.1 (IUCN 2012). We compared this to cropping areas in consensus land-cover maps (Tuanmu et al., 2014) and found that maps of Jung et al. (2020) more closely matched the concept of T7.1. | JR Ferrer-Paris, DA Keith |
| T7.2 Sown pastures and fields | ||
| T7.2.WM.nwx v1.0 | Mapping of intensive livestock pastures was based on fractional land use mapping (Ramankutty et al. 2008), dasymetric estimates of ruminant livestock density for cattle and sheep (Gilbert et al. 2010), and Human Appropriation of Net Primary Production (HANPP, Haberl et al. 2007). Fractional land use cover indicated firstly where pastures occur and secondly where they occupy a large portion of area relative to croplands. This helped to exclude intensive croplands that are also used to graze livestock, either through temporal rotation or on the margins of cropped paddocks (e.g. in south Asia). Livestock densities indicated where ruminants were important components of pasture systems, and helped exclude some rangelands with low livestock densities. Finally, HANPP helped exclude low productivity rangelands with high stocking rates and additional areas of cropland. We explored different combinations and thresholds for the input data layers, visually inspecting the output in South Asia, Australia, West Africa, and North and South America. We then mapped major occurrences where pasture area fraction greater than zero (PAF>0) and greater than cropland area fraction (PAF-CAF>0), densities of cattle or sheep were greater than 500 per cell, and 100 < HANPP < 700 gC/m²/yr. We examined the sensitivity of mapped area to variation in these thresholds and found no appreciatble change in the global mapped area when livestock density was varied by ±20% and marginal change in mapped area with variation in the other thresholds by the same amount. To represent this uncertainty, we mapped minor occurrences as the additional area where PAF>0, PAF-CAP>-0.2 and 80 < HANPP < 840 gC/m²/yr. We acknowledge significant untested assumptions and limitations on spatial predictors that challenge the global-scale delineation of pasture ecosystems with varied levels of human influence. We therefore advise appropriate caution in use of the spatial data for quantitative analysis. | JR Ferrer-Paris |
| T7.3 Plantations | ||
| T7.3.IM.alt v2.0 | Major occurrences of plantations were taken from the map of Habitat type 14.3 by Jung et al. (2020) based on the IUCN Habitats Classification Scheme v3.1 (IUCN 2012). We compared this to cropping areas in consensus land-cover maps (Tuanmu et al., 2014) and found that maps of Jung et al. (2020) more closely matched the concept of T7.3. Occurrences were extracted from fractional aggregated 1 km resolution base data (Jung et al. 2020), approximating 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| T7.3.web.alt v2.0 | Major occurrences of plantations were taken from the map of Habitat type 14.3 by Jung et al. (2020) based on the IUCN Habitats Classification Scheme v3.1 (IUCN 2012). We compared this to cropping areas in consensus land-cover maps (Tuanmu et al., 2014) and found that maps of Jung et al. (2020) more closely matched the concept of T7.3. | JR Ferrer-Paris, DA Keith |
| T7.4 Urban and industrial ecosystems | ||
| T7.4.WM.nwx v1.0 | The distribution of urban and industrial infrastructure lands was taken from a global land use/land cover map (LULC class 7 ‘built areas’) for the year 2020 at 10 metre resolution (Karra et al. 2021). Class 7 includes major road and rail networks, large homogenous impervious surfaces including parking structures, office buildings and residential housing, dense. Sparse villages may not be represented. We calculated the proportion of built area per square kilometre and applied a threshold of 1 to 5 % for minor occurrences and >5% for major occurrences. | JR Ferrer-Paris |
| T7.5 Derived semi-natural pastures and old fields | ||
| T7.5.WM.nwx v1.0 | We compared a land use dataset for the year 2000 with areas suitable for grazing (Erb et al. 2007). We selected areas that are a) suitable for grazing, b) predominantly used for grazing (grazing> forestry and grazing>cropland), but c) grazing use is intermediate (between 30-70%). We classified the best suitability areas (class 1 in Erb et al. 2007) as major occurrences, and the second best areas (class 2) as minor occurrences. | JR Ferrer-Paris |
| T7.5.WM.nwx v2.0 | We used land use and land suitability datasets for the year 2000 (Erb et al. 2007) and Human Appropriation of Net Primary Productivity (Haberl et al. 2007) to estimate the distribution of semi-natural pastures. We selected areas that were: a) suitable for livestock grazing; b) had a greater proportion of grazing lands than croplands or forestry lands (grazing> forestry and grazing>cropland); c) had intermediate to high cover of grazing lands (>30%); and intermediate to high estimates of Human Appropriation of Net Primary Productivity (20%| JR Ferrer-Paris, DA Keith |
|
| T7.5.web.mix v1.0 | Major and minor occurrences were initially identified using consensus land-cover maps (Tuanmu et al. 2014) and then cropped to selected terrestrial ecoregions (Dinerstein et al. 2017) at 30 arc seconds spatial resolution. Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| TF1.1 Tropical flooded forests and peat forests | ||
| TF1.1.IM.alt v4.0 | Major occurrences of tropical swamp forest and flooded forest were taken from the map of Habitat type 1.8 by Jung et al. (2020) based on the IUCN Habitats Classification Scheme v3.1 (IUCN 2012). We compared this to areas of of tropical swamp forest and flooded forest mapped Global Lakes and Wetlands Database (Lehner and Döll, 2004) as well as ecoregions with such forests mentioned in their description (Dinerstein et al., 2017), and found that maps of Jung et al. (2020) more closely matched the concept of TF1.1. Occurrences were extracted from fractional aggregated 1 km resolution base data (Jung et al. 2020), approximating 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| TF1.1.web.map v1.0 | Major occurrences of swamp and flooded forest were taken from the Global Lakes and Wetlands Database (Lehner and Döll 2004). Additional areas with minor occurrences identified in selected terrestrial ecoregions (Dinerstein et al. 2017). Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| TF1.2 Subtropical/temperate forested wetlands | ||
| TF1.2.IM.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| TF1.2.web.orig v2.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | JR Ferrer-Paris, DA Keith |
| TF1.3 Permanent marshes | ||
| TF1.3.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| TF1.3.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| TF1.4 Seasonal floodplain marshes | ||
| TF1.4.IM.mix v1.0 | Major occurrences of freshwater marshes and floodplains were taken from the Global Lakes and Wetlands Database (Lehner and Döll, 2004). Occurrences in boreal and polar climates were excluded by removing KoeppnGeiger_classes>26 in Beck et al., (2018). Additional areas with minor occurrences were identified in selected freshwater ecoregions (Abell et al., 2008). Ecoregions were selected if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Occurrences were aggregated to half degree spatial resolution. | JR Ferrer-Paris, DA Keith |
| TF1.4.web.map v1.0 | Major occurrences of freshwater marshes and floodplains were taken from the Global Lakes and Wetlands Database (Lehner and Döll 2004). Occurrences in boreal and polar climates were excluded by removing KoeppnGeiger_classes>26 in Beck et al. (2018). Additional areas with minor occurrences identified in selected freshwater ecoregions (Abell et al. 2008). Ecoregions were selected if: i) their descriptions mentioned features consistent with those identifed in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. | JR Ferrer-Paris, DA Keith |
| TF1.5 Episodic arid floodplains | ||
| TF1.5.IM.alt v2.0 | Locations of pan, brackish and saline wetlands were taken from the Global Lakes and Wetlands Database GLWD3 class 7 from Lehner and Döll (2004). Occurrences were aggregated to half degree spatial resolution. | JR Ferrer-Paris, DA Keith |
| TF1.5.WM.nwx v1.0 | Freshwater ecoregions (Abell et al., 2008) were identified as containing occurrences of these functional groups if: i) their descriptions mentioned features consistent with those identified in the profile of the Ecosystem Functional Group; and ii) if their location was consistent with the ecological drivers described in the profile. Within the selected ecoregions, remote sensing estimates of ephemeral surface water (classes 9 and 10 from Pekel et al., 2016) were used to identify major occurrences. The remaining areas of selected ecoregions were mapped as minor occurrences. | JR Ferrer-Paris, DA Keith |
| TF1.5.web.map v2.0 | Locations of pan, brackish and saline wetlands were taken from the Global Lakes and Wetlands Database GLWD3 class 7 from Lehner and Döll (2004). | JR Ferrer-Paris, DA Keith |
| TF1.6 Boreal, temperate and montane peat bogs | ||
| TF1.6.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| TF1.6.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| TF1.7 Boreal and temperate fens | ||
| TF1.7.IM.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al., 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently, they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |
| TF1.7.web.orig v1.0 | Terrestrial ecoregions containing major or minor occurrences of this ecosystem functional group were identified by consulting available ecoregion descriptions (Dinerstein et al. 2017), global and regional reviews, national and regional ecosystem maps, locations of relevant examples, and proofed by expert reviewers. Consequently they are coarse-scale indicative representations of distribution, except where they occupy small ecoregions. Ecoregions were mapped at 30 arc second spatial resolution. | DA Keith, JR Ferrer-Paris |