Integrating and scaling data to create ecosystem accounts relevant to management decisions 23rd London Group Meeting San Jose Costa Rica, 17- 20th October 2017 Dr. Heather Keith Research Fellow keather.keith@anu.edu.au Dr. Michael Vardon Visiting Fellow michael.vardon@anu.edu.au Emma Burns and David Lindenmayer

Questions to the London Group We defined ecosystem condition as time since disturbance. Has this been used previously or considered useful in future? In the current guidelines for ecosystem accounts, it appears that condition can be defined in two ways: (i) supply of ecosystem services, and (ii) ecosystem assets relative to a reference state. Should these two cases be described by different terms? A “sustainable yield” providing ecosystem services can be a subjective definition, and dependent on perspective (condition for nature or for human use). Should such an indicator be included in accounts that are an information system, and interpretation is performed at a later stage? Is there a difference between ecosystem extent and land cover extent? Using the same term in environmental accounts and ecosystem accounts would improve consistency and linkages between these accounts. Spatial units are related to classification systems and how available data can be integrated within these classifications. Available financial data were highly aggregated, so that land use could be linked only to the highest level of classification of ecosystem services (CICES). Are there suggestions about how to quantify lower levels of ecosystem services? Beyond accounting? We modelled changes in ecosystem services (net carbon stock change and water yield) in the counterfactual case – as if forest logging had not occurred. Is this considered a useful form of analysis to identify trade-offs between land use activities in their use of ecosystem services?

Central Highlands, Victoria, Australia Mountain Ash forest

Ecosystem condition related to time since disturbance Forest age – determined from stand-replacing events of wildfire and logging (high severity wildfire kills montane ash forest but not mixed species forest) years old regeneration period > 75 before 1939 56 – 75 1939 – 1959 33 – 55 1960 – 1982 7 – 32 1983 – 2008 0 – 6 2009 - 2015

Change in ecosystem extent and condition over time Time series Forest age class = ecosystem condition Land cover type: = ecosystem extent

Forest age influences water yield Reduction in water yield in younger montane ash Accounting for change over time in ecosystem condition requires ecosystem process functions in relation to a reference state, for example water yield and forest carbon dynamics.

Forest age influences carbon stocks Total biomass carbon stock changes over time due to rates of mortality, decomposition, regeneration, emissions, harvest and removal of wood, longevity of wood products, transfer and longevity in landfill. Wildfire Logging Initial biomass in old growth forest Ecosystem condition in terms of climate change mitigation value is the asset of total carbon stock (height up the Y axis) compared with the old growth forest (reference state). Rate of carbon sequestration (slope of the dark green regrowth forest curve) is not the correct metric for mitigation value. ‘Sustainable yield’ in terms of harvesting wood products is not equivalent to the maximum asset value that is considered the reference state for ecosystem condition.

Ecosystem condition for biodiversity 1. IUCN Red List of Threatened Ecosystems CRITERION A and B CRITERION C CRITERION D and E Ecological processes Assessment threat status and risk of loss of the Mountain Ash ecosystem. Threat status based on 5 criteria covering different spatial and temporal scales. Habitat attributes Disturbance factors Spatial extent and connectivity 8 species of aboreal marsupials Species dependencies

Ecosystem condition for biodiversity 2 Ecosystem condition for biodiversity 2. IUCN Red List of Threatened Species Ecosystem condition indicated by: Number of species listed Increase in severity of threat class over time

Abundance of arboreal marsupials declined over time Ecosystem condition for biodiversity 3. Animal abundance and habitat attributes Abundance of arboreal marsupials declined over time Number of arboreal marsupials positively related to hollow-bearing trees

Gains in non-monetary values Trade-offs defined explicitly Ceasing native timber harvesting increases ecosystem services for: Carbon sequestration and water provisioning calculated – Known gains Plantation timber provisioning and recreational services estimated – Potential gains Gains in non-monetary values

Defining spatial trade-offs Trade-offs defined spatially Defining spatial trade-offs Interaction index of ecosystem services of water, carbon and timber provisioning. ‘Hotspots’ are the highest values for all ecosystem services. Trade-offs are required when land uses conflict. Native forest available for harvesting Timber harvesting reduces ecosystem services for carbon storage and water yield

Questions to the London Group We defined ecosystem condition as time since disturbance. Has this been used previously or considered useful in future? In the current guidelines for ecosystem accounts, it appears that condition can be defined in two ways: (i) supply of ecosystem services, and (ii) ecosystem assets relative to a reference state. Should these two cases be described by different terms? A “sustainable yield” providing ecosystem services can be a subjective definition, and dependent on perspective (condition for nature or for human use). Should such an indicator be included in accounts that are an information system, and interpretation is performed at a later stage? Is there a difference between ecosystem extent and land cover extent? Using the same term in environmental accounts and ecosystem accounts would improve consistency and linkages between these accounts. Spatial units are related to classification systems and how available data can be integrated within these classifications. Available financial data were highly aggregated, so that land use could be linked only to the highest level of classification of ecosystem services (CICES). Are there suggestions about how to quantify lower levels of ecosystem services? Beyond accounting? We modelled changes in ecosystem services (net carbon stock change and water yield) in the counterfactual case – as if forest logging had not occurred. Is this considered a useful form of analysis to identify trade-offs between land use activities in their use of ecosystem services?