The operational map for Coastal Saltmarsh was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the eastern Regional Forest Agreements. The project’s Threatened Ecological Community Reference Panel (the Panel) reviewed the determination for Coastal Saltmarsh and agreed upon a set of diagnostic parameters for its identification. We identified that any treeless saline and sub saline native vegetation found in the intertidal zone had the potential to be Coastal Saltmarsh. We estimated the extent of the intertidal zone by using a fine scale digital elevation model to determine the highest astronomical tideline (HAT). We then mapped potential Coastal Saltmarsh by analysing recent fine scale three dimensional aerial imagery to identify any native vegetation that comprised of low-growing treeless communities and was located within the HAT and on the landward side of mangroves. Mapping criteria used a tree cover tolerance of up to 30% to include areas that had a mixed cover of mangrove, paperbark or casuarina species with a saltmarsh understorey. Exposed mudflats and banks were also mapped when they were visible. Our mapping covered 1.4 million hectares of State Forest within the south, central and north coast regions of NSW. We identified a total of 111.9 hectares of Coastal Saltmarsh within 14 State Forests along the east coast. The most extensive areas are located in Bermagui and Mogo State Forests on the south coast and in Wallaroo State Forests on the north coast. We validated our map of coastal saltmarsh using an existing independent map of estuarine habitats (Creese et al 2009). Our mapping consistently identified almost twice as much coastal saltmarsh as Creese et al (2009), but this was attributable to differences in the mapping criteria rather than any error. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.

Indicative map for Lowland Grassy Woodland: The indicative map for Lowland Grassy Woodland was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the eastern Regional Forest Agreements. The determination of Lowland Grassy Woodland was reviewed by the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel), and a set of diagnostic parameters for the identifying the Lowland Grassy Woodland TEC was agreed upon. Using these diagnostic parameters, we sampled candidate areas from existing vegetation maps to identify potential areas of Lowland Grassy Woodland occurrence in 296 000 hectares of State Forest and undertook additional mapping work using two independent mapping methods. Random Forest models (predictive habitat models) were generated using plot data and a selection of environmental variables. Aerial photo interpretation targeted stands of forests dominated by Eucalyptus tereticornis to refine the potential boundaries of Lowland Grassy Woodland. We tested whether Lowland Grassy Woodland was present in State Forest by completing systematic plot surveys within mapped areas indicating potential presence. We compared our collected data to a large regional pool of plot data that contained a subset of plots assigned to vegetation map units cited in the determination for the Lowland Grassy Woodland TEC (see Gellie 2005, Tozer et al 2006, and Keith and Bedward 1999). Our analysis of data confidently assigned only a few plots in State Forest to Lowland Grassy Woodland (2/43). From these results, we were unable to construct an operational map for Lowland Grassy Woodland. The relationship between the existing mapping cited in the determination and the plot data on State Forest was not strong enough to be a reliable basis for mapping the TEC. We also found that Eucalyptus tereticornis could not reliably be used as an indicator of Lowland Grassy Woodland in State forests. As a result, we were unable to map this TEC from the few confirmed sampling points without including a significant area of forest that was highly unlikely to be Lowland Grassy Woodland. However, we created indicative maps of Lowland Grassy Woodland by merging our predictive and API maps to provide an indication of the likely extent of Lowland Grassy Woodland in State Forests. Operational map for Brogo Wet Vine Forest: The operational map for Brogo Wet Wine Forest (BWVF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the eastern Regional Forest Agreements. We assessed whether BWVF was likely to be present in more than 296 000 hectares of State Forest in the South-east Corner Bioregion. The project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for BWVF and reaching an agreed interpretation of floristic, environmental and distributional characteristics. The Panel found that BWVF is primarily defined by a source vegetation community derived from quantitative floristic plot data (Keith and Bedward, 1999), with additional defining characteristics relating to bioregion and elevation. The Panel’s interpretation resulted in the identification of all State Forests located below an elevation threshold of 550 metres within the South East Corner Bioregion as potentially containing BWVF. We identified other potential areas of BWVF by overlaying the cited vegetation maps and any State Forest mapping where vegetation was dominated by or includes Eucalyptus tereticornis (a defining species of BWVF). Within these state forests, we used aerial photo interpretation (API) to identify and delineate potential areas of BWVF based on structural characteristics and overstorey and understorey attributes, namely dominance or inclusion of Eucalyptus tereticornis. We then compiled floristic plot data for all

The operational map for River-flat Eucalypt Forest (RFEF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The map was constructed in two parts, with State Forests to the north of Sydney being mapped in a separate process to those to the south of Sydney. We did this to minimise the risk that relationships between regional vegetation communities and the TEC would be confounded or masked by geographical variation or other major ecological gradients, which might otherwise be a significant risk if we had treated the full latitudinal range of the TEC as a single study area. In total, we assessed 1,218,000 hectares of State Forest across coastal NSW. This consisted of 868,000 hectares of State Forest on the north coast and more than 350,000 hectares of State Forest on the south coast. In both study areas, the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for RFEF and agreeing upon a set of diagnostic parameters for its identification. The Panel found that RFEF is primarily defined by floristic plot data and that it is mostly located on coastal floodplains and associated alluvial landforms. Following on from these conclusions, we started the mapping process by mapping the distribution of floodplains and alluvial soils and thus identifying possible areas of RFEF. For both the north and the south coast we used an existing map of coastal landforms and geology in combination with several fine-scale models of alluvial landform features to determine the likely extent of floodplains and alluvial soils within our study areas. We used aerial photograph interpretation (API) to assess the floristic and structural attributes of the vegetation cover found on our modelled alluvial environments, and thus delineated polygons likely to contain RFEF. We also used API to modify the boundaries of the modelled alluvial areas using a prescribed list of eucalypt, casuarina and melaleuca species in combination with the interpretation of landform elements relevant to alluvial and floodplain environments. We then compiled floristic plot data for all State Forest areas within our modelled alluvial landforms and API polygons. For both the north and the south coast the floristic plot data was sourced from both existing flora surveys held in the OEH VIS database and from targeted flora surveys conducted specifically for this project. We compared these plots with those previously assigned to flora communities listed in the determination of RFEF. Both dissimilarity-based methods and multivariate regression methods were used for the comparison. The results of the comparison were then used to assess the likelihood that the plots in State forests belonged to one or more of the communities listed in the RFEF determination. Following this, we developed a predictive statistical model of the probability of occurrence of RFEF using plot data and a selection of environmental and remote-sensing variables. For the north coast, we used a Random Forest model, while for the south coast we used a Boosted Regression Tree model. To create the operational map, we assigned every mapped API polygon to RFEF if appropriate based on the plot data, over-storey and understorey attributes, landform features and modelled probabilities underlying each API polygon. We mapped 3819 hectares of RFEF on the south coast and 198 hectares of RFEF on the north coast. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.

Operational map for Lowland Rainforest: The operational map for Lowland Rainforest (LORF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for LORF. The Panel found that the determination for LORF relies almost exclusively on a rainforest classification system described by Floyd (1990) where several rainforest ‘suballiances’ make up the LORF assemblage. Floyd’s suballiance classifications presented a challenge to our project as they were largely subjective and were not compatible with quantitative analysis, meaning that it was difficult to distinguish between the LORF TEC and other rainforest vegetation using statistically sound methods. To overcome some of these problems we revisited a set of reference sites that were assigned by Floyd to the suballiances cited in the LORF determination and in other rainforest TEC determinations, and collected new floristic data using standard flora survey methods. We also targeted a range of localities on State Forest that we considered likely to include LORF and other rainforest TECs based on the suballiance descriptions, cited localities in Floyd (1990), and preliminary distribution models. Over 300 new rainforest plots were combined with a large pool of existing data covering eastern NSW to construct a provisional revised rainforest classification. We used the rainforest groups derived from this analysis to compare the species composition of Floyd’s suballiances, determination assemblage lists and recent rainforest classifications included in regional classifications. Rainforest groups (and the plots that defined them) were assigned to the Floyd suballiance with the highest degree of floristic similarity. We conferred with the Panel to resolve any inconsistencies between the results of our analyses and statements relating to the distribution and composition of individual suballiances in Floyd (1990) and the determinations. We then used plot data and a selection of environmental and remote-sensing variables to develop a Random Forest (RF) model of the probability of occurrence of LORF. We assessed the location of plots assigned to LORF against the distribution of the RF model on and adjoining State Forests. We then completed detailed aerial photograph interpretation (API) using a prescribed set of mapping classes to delineate rainforest areas for a range of canopy cover thresholds. We constructed an operational map of LORF by assigning our API polygons as being LORF based on the modelled probabilities and plot data underlying the polygon. Our mapping identified a total of approximately 14,036 hectares of LORF, the vast majority of which was located in the north coast region. We mapped 13,209 hectares of LORF on the north coast, with the largest areas found in Ewingar and Unumgar State Forests. Only 827 hectares of LORF were mapped on the south coast, with the largest areas found in Yadboro and Currowan State Forests. Operational map for Lowland Rainforest on Floodplains: The operational map for Lowland Rainforest on Floodplains (LRFP) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for LRFP. The Panel found that the determination for LRFP relies mainly on a rainforest classification system described by Floyd (1990) where several rainforest ‘suballiances’ make up the LRFP assemblage. The determination also identifies a range of floodplain and alluvial descriptors. Floyd’s suballiance classifications presented a

The operational map for Tablelands Snow Gum, Black Sallee, Candlebark and Ribbon Gum Grassy Woodland (Tableland Snow Gum or TSG) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the eastern Regional Forest Agreements. The determination of TSG was reviewed by the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel), and a set of diagnostic parameters for identifying the TSG TEC was agreed upon. These parameters included the bioregions in which TSG is likely to be located in, landscape features such as elevation and geology, and quantitative floristic attributes from vegetation communities explicitly listed in the determination. Using these diagnostic parameters, we defined the study area as being all IBRA subregions that cover the 600-1400m elevation range within the South Eastern Highlands, Sydney Basin, South East Corner and Australian Alps bioregions. We then compiled floristic plot data for all State Forest areas within our study area. Floristic plot data was sourced from both existing flora surveys held in the OEH VIS database and from targeted flora surveys conducted specifically for this project. We compared these plots with those previously assigned to flora communities listed in the determination of TSG. Both dissimilarity-based methods and multivariate regression methods were used for the comparison. The results of the comparison were then used to assess the likelihood that the plots in State forests belonged to one or more of the communities listed in the TSG determination. We also conducted presence-absence predictive distribution modelling to identify potential distributions of each of the primary vegetation communities cited in the determination for TSG. The modelling predicted the likelihood of occurrence for each community across State Forests in the study area based on a modelled relationship with environmental and remotely sensed variables. As such, the modelling assisted in identifying un-surveyed areas of potential TSG habitat and guiding follow-up survey efforts and aerial photography interpretation (API) work. API assessment was carried out for all State Forests where the predictive modelling identified areas with high probability-of-occurrence values. We used recent high resolution stereo digital imagery in a digital 3D GIS environment to delineate areas of potential TSG based on observable patterns in canopy species dominance, understorey characteristics and landform elements. We constructed the operational map by assigning our API polygons as being TSG based on the extent to which the floristic plots within or near to each API polygon belonged to TSG. We used a precautionary approach and assessed a mapped polygon as TSG if the map unit to which it belonged contained any TSG plot. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.

Operational map for River-flat Eucalypt Forest: The operational map for River-flat Eucalypt Forest (RFEF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The map was constructed in two parts, with State Forests to the north of Sydney being mapped in a separate process to those to the south of Sydney. We did this to minimise the risk that relationships between regional vegetation communities and the TEC would be confounded or masked by geographical variation or other major ecological gradients, which might otherwise be a significant risk if we had treated the full latitudinal range of the TEC as a single study area. In total, we assessed 1,218,000 hectares of State Forest across coastal NSW. This consisted of 868,000 hectares of State Forest on the north coast and more than 350,000 hectares of State Forest on the south coast. In both study areas, the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for RFEF and agreeing upon a set of diagnostic parameters for its identification. The Panel found that RFEF is primarily defined by floristic plot data and that it is mostly located on coastal floodplains and associated alluvial landforms. Following on from these conclusions, we started the mapping process by mapping the distribution of floodplains and alluvial soils and thus identifying possible areas of RFEF. For both the north and the south coast we used an existing map of coastal landforms and geology in combination with several fine-scale models of alluvial landform features to determine the likely extent of floodplains and alluvial soils within our study areas. We used aerial photograph interpretation (API) to assess the floristic and structural attributes of the vegetation cover found on our modelled alluvial environments, and thus delineated polygons likely to contain RFEF. We also used API to modify the boundaries of the modelled alluvial areas using a prescribed list of eucalypt, casuarina and melaleuca species in combination with the interpretation of landform elements relevant to alluvial and floodplain environments. We then compiled floristic plot data for all State Forest areas within our modelled alluvial landforms and API polygons. For both the north and the south coast the floristic plot data was sourced from both existing flora surveys held in the OEH VIS database and from targeted flora surveys conducted specifically for this project. We compared these plots with those previously assigned to flora communities listed in the determination of RFEF. Both dissimilarity-based methods and multivariate regression methods were used for the comparison. The results of the comparison were then used to assess the likelihood that the plots in State forests belonged to one or more of the communities listed in the RFEF determination. Following this, we developed a predictive statistical model of the probability of occurrence of RFEF using plot data and a selection of environmental and remote-sensing variables. For the north coast, we used a Random Forest model, while for the south coast we used a Boosted Regression Tree model. To create the operational map, we assigned every mapped API polygon to RFEF if appropriate based on the plot data, over-storey and understorey attributes, landform features and modelled probabilities underlying each API polygon. We mapped 3819 hectares of RFEF on the south coast and 198 hectares of RFEF on the north coast. Operational map for Swamp Oak Floodplain Forest: The operational map for Swamp Oak Floodplain Forest (SOFF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The map was constructed in two parts, with

The operational map McKies Stringybark/Blackbutt Open Forest (MSBOF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the eastern Regional Forest Agreements. The project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) reviewed the determination for MSBOF and agreed upon a set of diagnostic parameters for its identification. These parameters included the presence of lateritic soils and the characteristic eucalypt species Eucalyptus mckieana and Eucalyptus andrewsii. Using these diagnostic parameters, we assessed whether MSBOF is present within more than 150 000 hectares of state forests within the Upper and Lower North East IFOA areas of the Nandewar and New England Bioregions. We began the assessment process by examining the distribution of Eucalyptus mckieana records in NSW Bionet and other available systematic plot data characterising regional vegetation patterns. We then used the results of a recently completed vegetation mapping in the region (OEH, 2015) to provide an indicative map of related vegetation communities within State Forests. From this initial investigation, we identified three state forests as candidate areas for MSBOF within the Project Study area – Mount Topper, New Valley and Clive State Forests. We collected systematic plot data from these state forests in mapped communities related to the TEC and from within stands of vegetation dominated by the primary eucalypt species described in the determination. On this basis we identified four plots within Mount Topper State Forest as being MSBOF TEC and went on to construct an operational map at a scale commensurate with the needs of forestry field operations. In total, we mapped a total of 201 hectares of MSBOF, 101 hectares of which were contained in the IFOA area. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.

The operational map for Swamp Oak Floodplain Forest (SOFF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The map was constructed in two parts, with State Forests to the north of Sydney being mapped in a separate process to those to the south of Sydney. We did this to minimise the risk that relationships between regional vegetation communities and the TEC would be confounded or masked by geographical variation or other major ecological gradients, which might otherwise be a significant risk if we had treated the full latitudinal range of the TEC as a single study area. In total, we assessed 1,218,000 hectares of State Forest across coastal NSW. This consisted of 868,000 hectares of State Forest on the north coast and more than 350,000 hectares of State Forest on the south coast. In both study areas, the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for SOFF and agreeing upon a set of diagnostic parameters for its identification. The Panel found that SOFF is primarily defined by floristic plot data and that it is mostly located on coastal floodplains and associated alluvial landforms. Following on from these conclusions, we started the mapping process by mapping the distribution of floodplains and alluvial soils and thus identifying possible areas of SOFF. For both the north and the south coast we used an existing map of coastal landforms and geology in combination with several fine-scale models of alluvial landform features to determine the likely extent of floodplains and alluvial soils within our study areas. We used aerial photograph interpretation (API) to assess floristic and structural attributes of the vegetation cover on our modelled alluvial environments, and thus delineated polygons likely to contain SOFF. We also used API to modify the boundaries of the modelled alluvial areas using a prescribed list of casuarina and melaleuca species in combination with the interpretation of landform elements relevant to alluvial and floodplain environments. We then compiled floristic plot data for all State Forest areas within our modelled alluvial landforms and API polygons. For both the north and the south coast the floristic plot data was sourced from both existing flora surveys held in the OEH VIS database and from targeted flora surveys conducted specifically for this project. We compared these plots with those previously assigned to flora communities listed in the determination of SOFF. Both dissimilarity-based methods and multivariate regression methods were used for the comparison. The results of the comparison were then used to assess the likelihood that the plots in State forests belonged to one or more of the communities listed in the SOFF determination. To create the operational map, we assigned every mapped API polygon to SOFF based on the plot data, over-storey and understorey attributes, landform features and model output underlying each API polygon. In total, we mapped approximately 272 hectares of SOFF across our full study area. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.

Environmental Impact Statement: Forest ecosystem classification and mapping for the Eden Comprehensive Regional Assessment : a report undertaken for the NSW CRA/RFA Steering Committee : project number NA 18EH

The operational map for Swamp Sclerophyll Forest (SSF) was constructed to resolve long-standing issues surrounding its identification, location and extent within the NSW State Forest estate covered by the coastal Integrated Forestry Operation Agreements. The map was constructed in two parts, with State Forests to the north of Sydney being mapped in a separate process to those to the south of Sydney. We did this to minimise the risk that relationships between regional vegetation communities and the TEC would be confounded or masked by geographical variation or other major ecological gradients, which might otherwise be a significant risk if we had treated the full latitudinal range of the TEC as a single study area. In total, we assessed 1,218,000 hectares of State Forest across coastal NSW. This consisted of 868,000 hectares of State Forest on the north coast and more than 350,000 hectares of State Forest on the south coast. In both study areas, the project’s Threatened Ecological Community (TEC) Reference Panel (the Panel) preceded the assessment process by reviewing the determination for SSF and agreeing upon a set of diagnostic parameters for its identification. The Panel found that SSF is primarily defined by floristic plot data and that it is mostly located on coastal floodplains and associated alluvial landforms. Following on from these conclusions, we started the mapping process by mapping the distribution of floodplains and alluvial soils and thus identifying possible areas of SSF. For both the north and the south coast we used an existing map of coastal landforms and geology in combination with several fine-scale models of alluvial landform features to determine the likely extent of floodplains and alluvial soils within our study areas. We used aerial photograph interpretation (API) to assess the floristic and structural attributes of the vegetation cover on our modelled alluvial environments, and thus delineated polygons likely to contain SSF. We also used API to modify the boundaries of the modelled alluvial areas using a prescribed list of eucalypt, casuarina and melaleuca species in combination with the interpretation of landform elements relevant to alluvial and floodplain environments. We then compiled floristic plot data for all State Forest areas within our modelled alluvial landforms and API polygons. For both the north and the south coast the floristic plot data was sourced from both existing flora surveys held in the OEH VIS database and from targeted flora surveys conducted specifically for this project. We compared these plots with those previously assigned to flora communities listed in the determination of SSF. Both dissimilarity-based methods and multivariate regression methods were used for the comparison. The results of the comparison were then used to assess the likelihood that the plots in State forests belonged to one or more of the communities listed in the SSF determination. Following this, we developed a predictive statistical model of the probability of occurrence of SSF using plot data and a selection of environmental and remote-sensing variables. For the north coast, we used a Random Forest model, while for the south coast we used a Boosted Regression Tree model. To create the operational map, we assigned every mapped API polygon to SSF if appropriate based on the plot data, over-storey and understorey attributes, landform features and modelled probabilities underlying each API polygon. In total, we mapped approximately 1131 hectares of SSF across out study area. Operational TEC Mapping have been derived by API at a viewing scale between 1-4000 using ADS40 50 cm pixel imagery and 1 m derived LIDAR DEM grids for floodplain EECs.