The power of seeds: the resilience of a restored plant community to fire

The post provided by Ebony Cowan

Smoke tent trail stimulating germination of the soil seedbank in restored Banksia woodlands. Photo credit: Ebony Cowan

This post refers to the article Soil seedbank development of smoke-responsive plant species in a 23-year restoration chronosequence and implications for resilience to fire by Cowan and colleagues published in Applied Vegetation Science (

Seeds are the building blocks for vegetation, and we found that after a fire disturbance, woodlands restored after mining can regenerate to a state similar pre-fire state, through relying on seeds! We studied soil seedbanks in restored Banksia woodlands near Perth, Western Australia. Banksia woodlands have evolved with fire and are fire-prone, so while many plant species are killed by fire, they can persist by developing soil seedbanks that respond to cues from fire, like smoke and heat.

To understand how the soil seedbank develops with time, instead of using the standard method of collecting soil and germinating seeds in a glasshouse, we smoked soil in the field and let germination occur under field conditions. By igniting plant material and letting smoke linger over soil in plots, we stimulated the germination of plant species typical of Banksia woodlands, allowing us to understand how the restored vegetation may respond to fire.

In fire-prone ecosystems, plant community development is usually influenced by time since fire. Some plants will resprout from existing plant material (rootstock), and others will regenerate from seeds. Next, plants develop resources to help them cope with the next fire, typically by developing resprouting buds or seedbanks. However, if it is too short or too long between one fire and the next, buds may not develop or begin to die, and/or seeds may not be available or viable. Determining the sweet spot for when fire may promote regeneration can be difficult in part because it varies among species.

Despite being rarely tested, we assume these processes occur in restoration, but unlike reference plant communities, restored communities often start from scratch. In our study, restoration includes application of a species-rich soil seedbanks to facilitate succession towards the target reference (Banksia woodland). We used smoke to test this assumption, and to assess soil seedbank development in vegetation aged between three and 26 years since the onset of restoration to determine how age influenced response to fire.

We went out every day for a month to count around 28,000 germinants after winter rains, most not taller than half of a matchstick! We recorded germination at every age, so regeneration from seeds after fire was possible regardless of age. The composition of the soil seedbanks changed with restoration age: younger ages were mostly composed of annual species that favour recently disturbed environments, while older ages were dominated by perennial species that can be long-lived and take a few years to develop seedbanks. For perennial species, which are most of the species in the ecosystem and are therefore important for biodiversity conservation and ecosystem functions, we found that their soil seedbank became more similar to the above-ground (established) vegetation in older ages. This suggests that fire in older ages will result in a similar vegetation community to what was there before burning.

Diverse soil seedbank that was composed of mostly perennials after smoking. We found 107 species across all restoration ages. Photo credit: Ebony Cowan

Measuring both above- and below-ground vegetation composition can provide valuable insights for predicting responses to disturbances. In a fire-prone environment, where a large proportion of species require smoke to germinate their seeds, smoke tents are a more informative method of exploring the soil seedbanks than a glasshouse. In a time of biodiversity loss, ensuring vegetation can respond to disturbances is important for its long-term persistence. Achieving this in restored ecosystems will promote biodiversity conservation.

Predicting responses to disturbances in restoration projects may be complicated. For ideas on how to do so, please see the below paper: Cowan, E.L., Standish, R.J., Miller, B.P., Enright, N.J. & Fontaine, J.B. (2021) A framework for measuring the effects of disturbance in restoration projects. Restoration Ecology, 29:e13379.

The authors acknowledge funding from the Holsworth Wildlife Research Endowment, the Australian Government and the Cooperative Research Centre for Transformations in Mining Economies, and support from Hanson Australia, City of Swan, and the Department of Biodiversity, Conservation and Attractions.

Smoke tent treatment to stimulate germination of seeds in the soil seedbank, many of which require smoke to germinate. We used this treatment to assess how the soil seedbank develops with time in restored banksia woodlands after sand mining. Video credit: Lauren Svejcar

Brief personal summary: Ebony Cowan studies the effects of disturbances on plant communities in disturbed ecosystems. She works with land managers to assist in the implementation of scientific research to promote biodiversity outcomes. She is undertaking her PhD at Murdoch University in Perth, Australia.