Viable seed dispersal after fire depends on highly variable flowering and cone production

Authors Robert Whelan and David Ayre provide further insight into their recently published Journal of Ecology article: Long inter‐fire intervals do not guarantee a large seed bank in a serotinous shrub (Banksia spinulosa Sm.)

Find out how fires influence the seedling recruitment of an Australian shrub. This research is especially poignant in the wake of the recent, devastating bushfires in Australia and also draws attention to the need for long-term ecological studies, investigating the ecological impact of fires on flora.


Now that the majority of the fires in the 2019-20 bushfire season in Australia have been extinguished, attention is being directed to predicting and enhancing the recovery of species and ecological processes, as the climate continues to change.

Fire burning through the shrubby understorey in eucalypt woodland, consuming the foliage of a Banksia spinulosa plant bearing cones from several previous years of flowering. The heat of the fire will open these cones, stimulating seed release. How many seeds will be viable, and how will the patterns of flowering and fruit set since the last fire affect the numbers of viable seeds released? Photo: Rob Whelan

These fires will have massive biological impacts and have already fuelled important debates about the extent and frequency of fuel-reduction burning as a fire management strategy. Many Australian forest tree and shrub species are considered resilient in the face of even high-intensity fire, partly because resprouting from epicormic buds or underground stems (lignotubers) is common and obvious. However, relatively few studies have investigated the effect of time since last fire on seeds available for recruitment.

In a phenomenon known as “serotiny”, many species protect seeds in tightly closed woody structures, similar to cones, that accumulate on the plant after each year’s flowering season and open only after being heated in a bushfire. The seeds are released onto a soil surface that has been made suitable for germination by the fire and will produce the next cohort of seedlings.

The serotinous “Hairpin Banksia”, Banksia spinulosa, showing the remains of an inflorescence produced in the current flowering season (left) and a ‘cone’, from the previous year’s flowering, with closed follicles that will protect the seeds from the heat of fire. Photo: John Tann – Wikimedia Commons.

The accumulation of cones on serotinous shrubs is so conspicuous that it would be easy to assume that the longer the time since the last fire the more seeds will be available for release when the next fire inevitably occurs. However, this assumption has not been empirically tested in many species.

Most of our knowledge comes from “snapshot studies”, which infer the age of the cones on a plant from indicators such as the location on the plant and the amount of weathering of remaining flower parts. Furthermore, the majority of these studies have focussed on so-called “obligate-seeder” species, which do not resprout after fire and whose future therefore depends solely on the seed bank.

Banksia spinulosa shrub resprouting from lignotuber 3 weeks after fire. Over 70% of the herb and shrub species in the dry sclerophyll forests of south-eastern New South Wales are resprouters with persistent seed banks. Photo: Rob Whelan

In this study we examined the accumulation of cones and seeds over twenty-years in a study of the resprouting shrub, Banksia spinulosa in south-eastern Australia. In 1986, we took over a study site at Barren Grounds Nature Reserve, near Wollongong, New South Wales, which had been established by our PhD student, Sue Carthew. Carthew had been measuring flowering and seed set in a random sample of about 50 plants of Banksia spinulosa within a large population, and we continued censusing flowering and fruit set annually.

Seven years later, we added a second population at Kioloa State Forest (taking advantage of the University of Wollongong’s annual Conservation Biology field camp for final-year undergraduate students) and started censusing over 300 additional plants. Each year, the students counted numbers of new inflorescences and counted and tagged the cones resulting from the previous year’s flowering. Each year we returned to the site a few weeks later to validate the data collected by the students.

After nearly 20 years for Barren Grounds and 13 years for Kioloa State Forest, all cones remaining on the plants were harvested, allowing a calculation of how many of the cumulated number produced over the years were “missing”. The harvested cones were heated to release the seeds and seed viability was tested in germination trials.

Banksia spinulosa cone “born” in the 1992 flowering season at Barren Grounds and still present on the plant at harvest thirteen years later. Photo: Rob Whelan

There were a few surprises…!

First, there was huge variation in annual flowering and cone production and an overall downward trend in flowering.

Second, there was enormous inter-plant variation in numbers of inflorescences and cones produced: 50% of the plants were responsible for producing almost all the cones throughout the study.

Third, less than half of the cones produced through the study were still remaining at the end of the study with many having been attacked by cockatoos in some years and seed predation by insect larvae caused further losses of stored seeds from intact cones.

Fourth, seed viability of remaining, intact seeds declined rapidly, meaning that the great majority of viable seeds were from the past few years of flowering and an average of only 2-4 viable seeds per plant would have been released had a fire occurred at the end of the study.

Yellow-tailed Black Cockatoo feeding on a recent cone of Banksia integrifolia. Seed predation by cockatoos and insect larvae can account for a large number of stored seeds. Photo: Rob Whelan

As a result of these processes, it seems that there is almost no prospect of a future fire coinciding with a substantial standing crop of viable seeds. We therefore predict very low seedling densities following the next fire.

One of the major wildfires that burned through eucalypt forests in south-eastern Australia this summer has apparently burned thorough our site at Kioloa State Forest. If the tagged plants can be located again, this presents an opportunity to test the prediction of poor seedling recruitment and to quantify mortality of established plants of this resprouting shrub species. Of the total area of eucalypt forest and woodland ecosystems in National Parks in the state of New South Wales, over a third has already burned this fire season.

R.J. Whelan & D.J. Ayre University of Wollongong, Australia


Read the full research article online: Long inter‐fire intervals do not guarantee a large seed bank in a serotinous shrub (Banksia spinulosa Sm.)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s