Project Description
Grassland Resurrection – Connecting Knowledge, Research & Management Forum
8 May, 2026. Fitzroy Town Hall , Melbourne, Victoria
Michael Goddard
Michael is a Senior Associate Botanist and Technical Lead at Biosis. He conducts flora and fauna surveys, ecological monitoring and impact assessments in a range of ecosystems, specialising in the Victorian Alps and Victorian Volcanic Plain bioregions. He has more than a decade of experience monitoring the condition and mapping the extent of threatened ecological communities, including Natural Temperate Grassland of the Victorian Volcanic Plain. Michael believes that sharing of knowledge and innovation will be crucial to arresting the decline in this critically endangered ecological community.
What will be required for Natural Temperate Grassland of the Victorian Volcanic Plain to avoid an extinction vortex?
Over the last 15 years, Biosis has periodically re-visited several grassland sites to assess the extent and condition of Natural Temperate Grassland of the Victorian Volcanic Plain (NTGVVP). While the location, size, land tenure and management regime of these sites vary, the trend for NTGVVP at these sites does not. NTGVVP has declined at all sites, including those managed for conservation. If these sites are representative of the bioregion, NTGVVP could be extinct in the wild by 2030. Significant investment in innovation, including robotics, machine learning and automation, will be needed to arrest this decline and avoid an extinction vortex.
Questions from the Forum
What is causing the decline in Natural Temperate Grassland that you have observed?
The decline in Natural Temperate Grassland that I have presented here is likely to have several causes, including weed invasion, excessive biomass accumulation, neglect, mismanagement and climate change.
Weed invasion
Weed invasion (often exacerbated by other factors) is the primary underlying cause of Natural Temperate Grassland decline. Native flora are being displaced by introduced flora (environmental weeds), leading to degradation in condition to the point that grasslands no longer meet the condition thresholds for Natural Temperate Grassland of the Victorian Volcanic Plain. As a starting point, a grassland must be dominated by native flora to be recognised as the Natural Temperate Grassland threatened ecological community (TSSC 2008). If weed invasion has progressed so far that introduced species now dominate the grassland, it will no longer meet the condition thresholds for Natural Temperate Grassland.
Importantly, many environmental weeds are still in the relatively early stages of their invasion and have not yet filled the full extent of their potential distribution as modelled by the Victorian Government in 2002 (KTRI 2002). Sleeper weeds include African Love-grass (perhaps better appreciated for its invasiveness in New South Wales), many Needle Grasses and Paspalum species. Recent results from various grassland monitoring projects that my colleagues and I undertake suggest that the threat of these weeds to Natural Temperate Grassland has worsened in recent years.
Excessive biomass accumulation
Many of the environmental weeds that are invading our native grasslands (e.g. Toowoomba Canary-grass, Chilean Needle-grass, African Love-grass etc.) have a relatively high proportion of their biomass above ground (compared with native species) and are often taller plants at maturity than many native species. As a result, these introduced species readily outcompete native species for resources, such as light. Over time, if there is no regular biomass control through burning, slashing/mowing and/or grazing, these introduced species ultimately displace the native species.
Neglect and mismanagement
I have observed that neglect and mismanagement of Natural Temperate Grassland has led to a decline in the extent and condition of Natural Temperate Grassland at a number of sites. Neglect generally comes about through a lack of knowledge, while mismanagement typically stems from a lack of resources.
For example, some land managers have a lack of awareness of environmental weeds and how to control them, while other land managers have never had the opportunity to learn the importance of the right fire, in the right place at the right time. In this respect, we have a lot to learn from Traditional Owners. Other land managers may be aware of weed threats and the importance of fire but simply lack the resources for weed control and burning e.g. a lack of time, money and/or availability of adequately trained personnel.
Climate change
Climate change is affecting the seasonality and amount of precipitation falling on the Victorian Volcanic Plain. Over the long term, climate change is likely to lead to an overall decline in annual rainfall, with reduced rainfall in autumn, winter and spring but increased rainfall in summer (DELWP 2019).
The seasonality of rainfall is important to the germination and growth of grassland flora. Decreased cool season rainfall and increased summer rainfall may inhibit the growth of native flora (e.g. native forbs) and promote the growth of environmental weeds (Scott and Morgan 2012).
2020, 2021 and 2022 were all wet years, with a La Niña weather pattern and above average rainfall for the Victorian Volcanic Plain. Increased rainfall is known to increase the cover of introduced flora in grasslands of the Victorian Volcanic Plain (Zeeman 2017).
It is unknown whether the recent proliferation of environmental weeds that has contributed to a decline in Natural Temperate Grassland at my case study sites would be reversed by consecutive years of drought or whether this is the ‘new normal’, particularly given the predicted effects of climate change on rainfall patterns.
All of the above
It is likely that the cumulative effect of a combination of the above factors has contributed to the weed invasion observed at many case study sites that I have presented. Moreover, it is possible that a positive feedback loop has exacerbated recent observed declines in Natural Temperate Grassland.
When the integrity of a native grassland remains intact, it is known to more strongly resist weed invasion. For example, an intact soil crust inhibits the germination of non-native species (Morgan 2006). Many years of weed invasion will have contributed to fragmentation of native grassland and a decline in overall ecological integrity at the case study sites, which is likely to have left the native grasslands vulnerable to further weed invasion.
Importantly, the case studies that I have presented show decline due to functional loss and loss of condition, not what I refer to as complete loss due to irreversible clearing. While it is sobering to think that these case studies do not capture the full range of grassland loss that is occurring, it is reassuring to think that the loss presented in these case studies is theoretically reversible with some concerted restoration effort.
What mapping software would you recommend?
QGIS and QField are free and open-source software packages that offer the functionality needed to map grasslands.
If most grassland sites are in decline, what types of sites should we be prioritising for restoration?
In my mind, there are three types of Natural Temperate Grassland restoration, all of which should be prioritised to some extent. Each type of restoration site has its place.
(1) Existing sites that still support Natural Temperate Grassland
We need to be finding what is left and actually managing it properly. While this might only slow the decline for now, it will at least buy us some time to develop some of the innovations and gather some of the resources that we will need to effectively and efficiently manage and restore grasslands in the long-term.
(2) Recently degraded sites
We need to be finding sites that are not yet too far gone. These sites probably offer good potential for large scale restoration, possibly without the need for more intensive restoration methods like soil scalping. These types of sites probably represent some of the most achievable broad-scale gains.
(3) Sites where Natural Temperate Grassland disappeared long ago
Small-scale urban restorations (like the City of Melbourne’s Wildflower Meadows project along Gatehouse Street in Royal Park or the recent median-strip grassland that Yarra City Council has planted along Michael Street in Fitzroy North) would serve a double-purpose.
They not only restore grasslands to areas where they might have once occurred long ago but they would literally put grasslands on our doorsteps, so that more people appreciate the significance and beauty of our native grasslands. Without broader appreciation of grasslands, we will struggle to generate the interest and resources needed to restore them
How long until the wheeled drones take my job?
I would be surprised if a drone or robot took the job of a grassland restoration ecologist within my lifetime. I hope that my presentation highlighted the urgency with which we need to be restoring grasslands, but also the lack of resources that we currently have (e.g. even grasslands in conservation sites are in decline).
We need more restoration ecologists than ever before, but we cannot wait for more restoration ecologists to be trained – it will be too little, too late. We will need an army of restoration ecologists, complemented by the adoption and adaptation of new tools.
Drones and robots will simply be another tool for restoration ecologists to use to maximise the effectiveness and efficiency of their restoration efforts. For example, there are often very limited windows of time within a day, month or season for undertaking grassland weed management. There may only be a short period of time to tackle a particular weed species before it sets seed and this period of time shortens even further when factoring in the need for suitable environmental conditions for weed management (e.g. wind, rainfall and temperature must be appropriate for herbicide application). We cannot possibly expect our current workforce of highly trained and qualified ecologists to drop everything (e.g. seed collection, propagation, biomass management etc.) and treat this particular weed species during the narrow window in which it can be treated and across all the grassland sites that it has invaded. Restoration ecologists cannot be everywhere all at once! Sometimes simply getting personnel out on site in time is prohibitive. We do not have the luxury of having people sitting idle on site, awaiting suitable conditions for weed control.
In contrast, robots could be permanently stationed on grassland sites and ready to go at a moment’s notice. When conditions are not appropriate for weed control, they could be roaming the grasslands and mapping weeds for future control. When the conditions are right, they could then return to mapped weed infestations and treat them. In my mind, these robots would have docking stations on site, where they recharge and collect different management payloads e.g. a mechanical weed removal tool, a flame weeder, a herbicide applicator, a seed distributor.
Ultimately, humans (grassland restoration ecologists) will still be needed to train, guide and supervise these machines, re-stock their raw materials, monitor their progress and adapt their management actions accordingly. Robots would undertake some of the more mundane but time-critical tasks (e.g. spot spraying high threat weeds), freeing up restoration ecologists to focus their skills, expertise and attention on restoration challenges that they might not otherwise have time to tackle (e.g. ecological burns or seed production/collection). We need to be throwing every tool we have and every tool we can develop at grassland resurrection. The challenge is so great that no grassland restoration ecologist should ever be out of work.
Offsets are generally crap and underwhelming…. How do you mitigate the shitness of offsets ?
The 2020 Samuel Review of the EPBC Act highlighted clear deficiencies in the EPBC Act Environmental Offset Policy and implementation of environmental offsets. The Samuel Review proposed a number of improvements, which are the subject of current draft reforms to the EPBC Act (e.g. the National Environmental Standard for Environmental Offsets).
In my opinion, there are three main measures to mitigate against the deficiencies in offsets that the Samuel Review highlighted:
(1) Avoidance and minimisation
Offsets should only be a last resort. There must be a true and genuine effort to avoid impacts to Natural Temperate Grassland.
(2) Repair and rehabilitation
For impacts that are unavoidable, there should be a commitment to repair those impacts where possible and rehabilitate Natural Temperate Grassland that is affected.
(3) Restoration offsets
Where offsets are required to compensate for unavoidable impacts, there should be some acknowledgement that not all offsets are created equal. If we are to value the long-term viability of Natural Temperate Grassland, restoration offsets are clearly preferable to ‘averted loss’ offsets. If successful, restoration offsets provide a clear gain in area and quality of grassland to compensate for loss elsewhere.
We need to be incentivising restoration offsets and, in doing so, incentivising the development of some of the solutions that will be needed for Natural Temperate Grassland to avoid an extinction vortex. Incentivising restoration offsets would promote a stronger market for grassland restoration, provide more certainty to grassland seed producers and promote innovation in grassland restoration.
Ultimately, offsets do not need to be underwhelming. They can be part of the solution.
