Monitoring impacting events - fire, rabbits, rainfall, seasonality

Our research aim at the Australian Botanic Garden, Mount Annan has been to understand the ecology of the woodland as it recovers from nearly two centuries of rural use, and apply this knowledge to conservation management. Our main performance indicator has been the ongoing survival and expansion of all the native plant species present in the woodland, while reducing the abundance of exotic species. Treatments used have been ecological burning, exclosure of herbivores and weed eradication. No plantings of any species have been made.

• Some effects of fire
• Rabbit impacts - exclosure plot observations
• Notes from the 2002 drought
• Importance of rainfall
• Seasonality - annual and short-term changes - 2001-2002
• Value of long-term monitoring

Some effects of fire

There were no signs of any recent burning in our woodland in 1988, and the area had evidently remained unburnt at least since the early 1980s, and probably much longer, as stock grazing presumably kept the grass low. However there was vigorous regrowth taking place in the early years of our recording, and little evidence of aging or senescent shrubs. Aging or senescent shrubs are often conspicuous in Sydney Sandstone vegetation that has remained unburnt for long periods.

However we assumed that fire would be important in Cumberland Plain Woodland ecology, and did want to try out treatments that might remove weed species and benefit native species. With this as an aim we burnt part of the woodland in September 1991.

Three plots were burnt in the September 1991 fire. Average native plant species richness dropped slightly immediately after the fire, and then returned to prefire levels (see Figure below). However native species richness at unburnt plots showed similar drops, indicating that this was weather-related rather than fire-related (1992 and 1993 had below average rainfall). Indeed though there was a difference in species richness between burnt and unburnt sites before the fire, these differences were reduced after the fire.

Average exotic species richness increased significantly immediately after the fire by over 50% compared with prefire levels 3 years earlier (see figure below), but in the next year it dropped to levels similar to that of unburnt sites and remained similar to them for the next 8 years.

In general we did not note any obvious long-term changes in native species composition resulting from the burn and within 3 years the burned area was not noticeably different from adjacent unburnt areas.The main result of the 1991 fire however was that we were able to document the fire responses for a large number of species and demonstrate that the majority of these species could resprout after fire.

Ten years later in September 2001 we burnt the same area again, and the following year September 2002 we burnt an adjacent area. Both fires were again Spring burns and of low intensity.

Fire has a different dynamic in Cumberland Plain Woodland compared with shrubby woodland vegetation on sandstone. Differences in fuel loads between the two are crucial. In our woodland there is not the heavy fuel build-up that comes from the characteristic shrub-dominated component in woodland on sandstone. In Cumberland Plain Woodland most sticks and persistent leaves come from the widely-spaced eucalypts, shrubs are sparse and make little contribution, and herbs and grasses have soft leaves that decompose faster than the hard leaves of sandstone-growing shrubs. Thus in Cumberland Plain Woodland, fires move rapidly through the groundlayer when it is dry enough, but do not generate high temperatures. These observations were confirmed in our 2005 fires where fires in long unburnt woodland were more patchy and of lower intensity than in adjacent areas of grassland.

A surprise from the fires

A major finding from the fires surprised us. Immediate responses were slow because of dry conditions, but we discovered that most species responded to subsequent rain in the bare space created by fire, rather than to the direct effects of fire - heat, smoke or ash. We found almost as many different types of seedlings emerging in bare spaces in unburnt areas after rain, as came up in the burnt areas. 

Woodland fire September 2002

Grassland after fire September 2005 

Map of Mount Annan Botanic Garden Conservation area burns.

• Pale pink 1991 & 2001
• Pink 2001
• Red 2002
• Orange 2005

Seedling germination after fire and significant rainfall

Species - Natives

Burnt

Unburnt,

before rain

after rain

after rain

Acacia implexa?

X

Chloris ventricosa?

X

Cotula australis

X

X

Cymbonotus lawsonianus

X

X

Cyperus gracilis

XX

Desmodium brachypodum

X

X

Desmodium varians?

X

Dichondra repens

XX

XX

XX

Dodonaea viscosa

X

X

Einadia hastata

XX

XX

Einadia nutans

XX

XX

Eremophila debilis

XX

X

Erodium crinitum

X

Eucalyptus crebra?

X

Euchiton sp.

X

X

Fimbristylis dichotoma

X

Galium propinquum

X

Geranium homeanum

X

X

Glycine tabacina

XX

X

Hypericum gramineum?

X

Mentha satureioides

X

Microlaena stipoides?

X

Oxalis perennans

XX

XX

Phyllanthus virgatus

X

Plectranthus parviflorus

X

X

Poranthera microphylla

X

Sida corrugata

XX

X

Solanum prinophyllum

XX

XX

Solenogyne dominii

X

Wahlenbergia gracilis

X

X

TOTAL NATIVES 30

3

26

18

Species - Exotics

Burnt

Unburnt

Before rain

After rain

After rain

*Anagallis arvensis

XX

XX

*Ciclospermum leptophyllum

X

X

*Cirsium vulgare

X

X

X

*Gomphocarpus fruticosus

X

*Heliotropium amplexicaule

XX

XX

*Hypericum perforatum

X

*Hypochaeris microcephala

X

*Hypochaeris radicata

X

*Medicago lupulina?

XX

*Modiola caroliniana

X

X

*Phytolacca indica

X

X

*Plantago lanceolata

XX

X

*Senecio madagascariensis

XX

XX

*Setaria gracilis

X

*Sida rhombifolia

XX

XX

XX

*Solanum nigrum

X

XX

XX

*Sonchus oleraceus

X

X

*Trifolium (incl. 3 species)

XX

XX

*Verbena bonariensis

X

TOTAL EXOTICS 18-21

4-5

16-19

14-17

Unidentified dicotyledons

3 species

Unidentified monocotyledons

3+species

Notes from the 2002 drought

The drought in 2002 occurred at just the right time in our recording sequence. Our monthly recording had begun in 2001 and had given us a year of average year effects. This was followed by the drought year - there was virtually no rain from the end of March to December 2002. Some rain came in late Summer and Autumn 2003, with really good rain in May. (Rainfall 2001: 631 mm, 2002: 524 mm, 2003: 644 mm, longterm average 828 mm)

• Native species we had classified as annuals - and considered lost as we had not seen them since 1988 - reappeared! This was really exciting! These species were behaving, not as conventional annuals, that go through their life cycle regularly every year, but like ephemerals, species in the semi-arid areas of Australia that only appear every few years after good rainfall.
  
• Daucus glochidiatus, a native species in the carrot family, reappeared patchily, but was abundant in those patches. Most plants emerged in areas that hadn’t been burnt. We couldn’t have missed it if it had been present in other years. It is a herb that grows to 50 cm tall in our woodland, and, like a carrot, produces small fruits covered in adhesive hairs.
  
• Ranunculus sessiliflorus, an even smaller herb up to 25 cm high, reappeared widely throughout the woodland, in both burnt and unburnt areas. It is in the buttercup family, and, though its flowers are tiny, it flowered profusely.
  
• Both are small plants that don’t need a lot of resources to grow to maturity and produce seed for the next generation.
  
• Also interestingly, these two species appeared at different times after the breaking of the drought. The Ranunculus sessiliflorus appeared in May, while the Daucus glochidiatus was not seen until October, so the two species have different dormancy characteristics in their seeds. Even though they both reappeared after drought and rain, they did it differently in relation to the seasons. This is an example of how plants behave in a similar fashion - but in slightly different ways. We believe this is a factor that enables coexistence of many species together, as they are not competing directly with each other.
  
• A third species behaving as an ephemeral appeared for the first time, a tiny herb in the chenopod family, barely 5 cm tall - Chenopodium carinatum. We found only a few of these tiny plants, so they could easily have been missed before. Being tiny, they completed their life cycle very quickly after the rain, flowering in May and setting seed in June.
  
• Ironically, although it was the drought that seemed to prepare these species for germination, ephemerals need periodic episodes of high rainfall, like that in May, to enable production of good seed crops to replenish their soil seedbanks.

^ back to top

The importance of rainfall

While we were well aware that rainfall was important for the successful growth of woodland species we had not truly realised its importance in the initiation of growth and particularly seedling growth events. We had always assumed that rainfall was distributed more or less throughout the year and was not generally a limiting factor, rather that triggering factors like fire or increasing temperatures were most important in regulating growth and recruitment. Rainfall in the late 1980s and early 1990s was generally average or above. However since 2000 there have been a series of dry and sometimes hot years. 2002 was a particularly dry and hot year, with record high temperatures in December.

In the area burnt in our 2002 burn there was little regeneration because of the lack of rainfall, so in summer 2002-2003 we took soil samples and watered them in the glasshouse to see what might be in the soil seedbank. What came up in the soil samples was an impressive array of seedlings many of which we had never seen in the field, though we were familiar with the adult plants there. We also got seedlings of one species Ranunculus sessiliflorus, which we had recorded once back in 1990, and, but for the fact that we had kept a small specimen of it, we would have regarded as an error of identification. As it was, we regarded the species as likely to be extinct in the woodland. Subsequent rain in March 2003 (drought-breaking) resulted in the germination of many seedlings in the woodland, including Ranunculus sessiliflorus and Daucus glochidiatus which has also been recorded only once before. Seedlings of both species, and many others, were recorded in both burnt and unburnt areas, indicating that it was particular site conditions not the effect of fire that allowed seedling establishment, though the open conditions in some burnt areas may have helped by providing locally higher soil temperatures or reduced competition from established plants.

Table

Number of woodland species with seedlings recorded in burnt (September 2001) or unburnt sites before and/or after good rain (February-March 2002), the Australian Botanic Garden, Mount Annan .

Seedlings observed October 2001-August 2002

Short-lived ephemeral species

We have been able to compile a list of species, almost all short-lived ephemerals, that recruited well in drought-breaking conditions, ie substantial rain following a long hot dry period. The list includes Ranunculus sessiliflorus, Daucus glochidiatus,*Conyza sumatrensis. While some of these species will occur at other times, generally in smaller numbers, we suggest that factors relating to drought conditions such as temperature may play a part in their dormancy breaking mechanisms. Further work is needed.

As a result of these observations we are convinced that the main force driving processes in the woodland is rainfall rather than fire or temperature. The decreased rainfall events expected as climate change takes hold may have a significant impact on the long term floristic composition of the woodland.

^ back to top

Rain

Seedlings of *Sida rhombifolia

Seasonality - annual and short-term changes 2001-2002

Any patch of ground in Cumberland Plain Woodland is likely to change the way it looks during the year. The herbs and grasses grow during favourable weather and die back when it’s hot and dry, or when it’s too cold. In recording species presence in our plots, the number of species has varied from as low as 2 per square metre during a very dry time to 27 after rain has stimulated growth. The number of species one can find varies over space as well as time. In our 5 m x 5 m plots, the total number of species recorded as present at some time has varied from 34 to 53.

Changes over a two year period 2001-2002, based on individual species abundance (frequency) from the monthly recording of Plots 4 and 9, show the following features:

• Most species fluctuate in abundance showing a peak in frequency between March and August (autumn-winter) with a decrease to a low frequency in December- January. This follows the mean annual rainfall trend which peaks in January with a trough in August- September. This is followed by a period of high maximum temperatures in November-March which slows recovery.
  
• Rainfall in 2001-2002 was below average in both years (Campbelltown mean annual rainfall 747 mm (78 years) Actual rainfall Camden 2001-530 mm, 2002-459 mm) though most rain did fall within the March to August period.
  
• Many species disappeared completely during the dry period, to reappear in March, not just the annuals e.g. *Anagallis arvensis, but perennials including Caesia parviflora, Geranium homeanum, Hypoxis hygrometrica, Wahlenbergia gracilis, Tricoryne elatior, Sporobolus creber. Even more species disappeared in the extreme 2002-2003 summer.
  
• A few species did not show a seasonal fluctuation during 2001-2002 e.g. Brunoniella australis, Dichondra repens, Sida corrugata (though they showed in 2003), Themeda australis, Scleria mackaviensis (fluctuating but in different way).
  
• Spring favouring species are few, Lomandra filiformis shows a later growth period June to November (spring) than others such the similar-sized Cyperus gracilis February-July(autumn-winter)
  
• Some species are opportunists - The single high rainfall event in February 2002 (250 mm), triggered an immediate clearly marked growth/flowering response in Tricoryne elatior, and also Bothriochloa macra, but was absorbed as part of the autumn seasonal growth by other species.
  
• Some species grow similarly in the two years, Chloris ventricosa, Cyperus gracilis, Oxalis perennans, Sporobolus creber. Some species grew better in 2001 e.g. Veronica plebeia, Hypericum gramineum, while others grew better in 2002 *Senecio madagascariensis, Geranium homeanum, Bothriochloa macra. 2002 had a lower rainfall but a lot in February.
  
• Winter cold did not have a noticeable affect on growth patterns. Sida corrugata showed a decrease in August 2001 but not in 2002. *Sida rhombifolia shows a decrease in abundance in July of both years but decreases mainly in dry summers.
  
• In terms of weed native interactions some weeds deteriorate over the period e.g. *Paspalum dilatatum, *Verbena bonariensis but others are expanding e.g. *Hypericum perforatum
  
• Species tended to show similar patterns at the different sites even though abundance differed e.g. Chloris ventricosa, Brunoniella australis. For some species there were some differences between the two sites, e.g. Bursaria spinosa fluctuated seasonally at Plot 4 but remained constant at Plot 9. For Wahlenbergia gracilis abundance was higher in 2001 than in 2002 at Plot 9, but higher in 2002 compared with 2001 at Plot 4.

Our picture of the Cumberland Plain Woodland flora at the Australian Botanic Garden is of a flora of species with characteristics that enable persistence in the face of fluctuating seasonal rainfall and temperature conditions. Most species are perennial and able to survive severe dry periods at ground level as rootstocks or soil stored seed. These characteristics also allow them to survive fire events. Growth cycles are aligned with longterm annual rainfall patterns, with most species being autumn-winter growing and flowering. Under good rainfall conditions growing and flowering will extend into spring with a slight mid Winter hiatus in very cold years. This is different to the Sandstone flora where spring flowering and growth is the main response.

Some difficulties in monitoring Cumberland Plain  Woodland

• Dry conditions seasonally and even daily
• Tiny species make location and identification difficult, particularly if they are also short lived
• Similar closely-related species

^ back to top

Removal of African Olive by cutting and poisoning

 
Bush regeneration volunteers