This appendix provides relevant information about the known and potential threats to the Banksia Woodlands ecological community. The Threatened Species Scientific Committee will be using available data to assess the impacts of these threats on the Banksia Woodlands ecological community. Specifically, these impacts will be addressed against listing criteria, to determine eligibility for listing against each criterion. The six listing criteria are outlined in the Guidelines for nominating and assessing threatened ecological communities.
Land clearing and impacts associated with fragmentation
Land clearing, development and intensification of land use, results in habitat loss, fragmentation and modification (DPaW, 2014). Clearing reduces the extent of the ecological community and exacerbates patch isolation, reducing connectivity between remnants. Connectivity between remnants of the ecological community and other native vegetation is an important determinant of habitat quality at the landscape scale for native flora and fauna as well as for overall condition and persistence of the ecological community.
Urbanisation has been the main driver of Banksia Woodland fragmentation, starting shortly after Perth was founded in 1829. Urban growth has been most intense since the 1960s, largely driven by a mining boom, and the population is estimated to reach 3.5 million by 2050, which is an increase of almost 70 per cent on the 2015 population (Weller, 2009; Ramalho et al., 2014; Government of Western Australia, 2015). Banksia Woodlands in Perth and surrounds persist in a few large conservation and Crown Land areas on the current city boundaries, and in urban reserves (most of which are small and isolated), linear strips on roadside verges, and rural private properties (Ramalho et al., 2014).
Fragmentation results in reduced connectivity of the floral and faunal components of the ecological community. It can impede movement and dispersal of plants and animals, especially where unsuitable habitat may separate fragments. Patches in fragmented landscapes also have greater levels of ‘edge effects’ such as human disturbance, weed invasion and feral animal impact than larger, more connected patches due the greater patch edge to area ratio. In narrow remnants, where the edge to area ratio is larger, it is easier for disturbances to invade relatively further into patches and impact on the ‘core’ of the patch.
Many Banksia spp. require the co-incidence of burnt occupied and unoccupied sites to allow seed dispersal and colonization to occur (Cowling and Lamont, 1987; Cowling et al., 1987, 1990; Enright et al., 1998a,b; Groom and Lamont, 2015). Fragmentation creates barriers for plant (Banksia) dispersal and colonisation where land in between remnants is primarily urban or used for intensive agriculture. Consequently there are fewer opportunities for colonisation due to rare long-distance dispersal events, which are required to adapt to rapid climate change (Yates et al., 2010).
Fragmentation impacts may take some time to become evident, however are more rapid in smaller remnants. Ramalho et al. (2014) found that richness of native herbaceous species in Banksia Woodlands declined with time since isolation, mainly in the smaller remnants, and this was associated with altered soil properties. In small remnants the native plant species richness in small remnants halved in only a few decades after isolation. Furthermore, increased litter depth (possibly indicating higher productivity) and increased abundance of non-native herbaceous species in the older and smaller remnants was associated with a decline in the abundance of native herbaceous species (Ramalho et al., 2014).
Long-term climate variability is affecting the southwest of Western Australia, which is experiencing a trend of increasing temperatures and declining rainfall. The number of days per year hotter than 40°C has been increasing since the 1990s, and late autumn and winter rainfall (the period of most importance for native plant growth in this region) has been decreasing (CSIRO, 2012; DPaW, 2014). Since 1970 there has been a 17 per cent decline in average winter rainfall in the southwest of Australia. The decline in this region has also been characterised by a lack of very wet winters. The reduction in rainfall is amplified as decreased streamflow in rivers and streams. In the far southwest, streamflow has declined by more than 50 percent since the mid-1970s (CSIRO, 2014). This is having an impact on plant reproduction and seedling recruitment (Keith et al. 2014).
Further decreases in average rainfall are expected over southwest Western Australia compared with the climate of 1980 to 1999. Based on modelling of carbon emission scenarios, a zero to 20 per cent decrease is expected by 2070 for low emissions with a 30 per cent decrease to 5 per cent increase by 2070 for high emissions, with largest decreases in winter and spring (CSIRO, 2014).
Urban heat islands can affect local climates and have effects on nearby remnants. Urban heat islands are caused by three factors in urban development - built materials trapping heat, urban machinery producing waste heat and the removal of trees and vegetation (and their associated shading and cooling functions) (Brown et al., 2013).
One of the most significant threats to wetland and woodland ecosystems in the Swan Coastal Plain is the reduction of groundwater levels as a result of an increase in groundwater abstraction (including production bores), patterns in water regulation and decreased rainfall and subsequent recharge to the groundwater system. The dominant, deep-rooted Banksia species of the ecological community are considered to be groundwater dependant and are therefore particularly susceptible to impacts from groundwater drawdown (Canham et al., 2009). Impacts related to groundwater drawdown range from a gradual change in the structure and composition of the ecological community to sudden and widespread vegetation death (Groom et al., 2000).
Risk to Banksia Woodlands depends on the floristic community type present and its corresponding dependence on groundwater resources in the region (Groves, 2014). Previous studies comparing Banksia woodlands where groundwater extraction was occurring to those in unaffected areas show that deep-rooted tree and shrub species are more susceptible to water and temperature stress than shallow-rooted species (Groom et al., 2000; Horwitz et al., 2009). The high degree of groundwater dependence makes Banksias highly vulnerable to rapid changes in water table elevation, and historically, large numbers of Banksias have died near water supply production bores due to rapid water table decline caused by groundwater pumping during exceptionally hot summers (Groom et al., 2000).
Where impacts of decreased groundwater availability on the ecological community result in a change in plant composition and structure, there is a shift in plant community composition from phreatophytic to vadophytic species (or deep-rooted to shallow-rooted) as groundwater resources become unavailable (Groves, 2014).
Groundwater decline is not only influenced by extraction but also by declining recharge/rainfall rates as a result of climate change. Climate data has shown a decrease in average annual rainfall since 1970, dominated by reduced winter rainfall (CSIRO, 2014). Climate change may also result in temporal and spatial changes in hydrology within the Swan Coastal Plain. Changes in soil temperature and distribution of surface water as a result of a warming climate may have implications for Banksia species that are restricted to lower-lying areas, such as B. ilicifolia (Groves, 2014). Those species restricted to waterlogged areas rely heavily on subsurface soil moisture and groundwater during periods of summer drought. Climate change may reduce the number of seasonally waterlogged areas, as well as increasing the depth to groundwater, resulting in a decrease in the number of phreatophytic species in these areas (Groom, 2004).
Groundwater acidification and related effects
Groundwater decline may also result in flow on effects due to decreased access to the water table, which can impact fauna species dependent on seasonal wetlands (e.g. amphibians; Mitchell et al. 2013)
Banksias are susceptible to death or decline due to increased acidity and soluble aluminium concentrations in subsoil porewater in areas of rapid decline of the water table in areas underlain by Bassendean dunes. Soils in these areas have a low buffering capacity and are known to contain sufficient pyrite to create acid sulfate soil conditions when the water table declines (Prakongkep et al., 2012).
Prior to European settlement, some fires occurred through lightning strikes and Aboriginal burning of the landscape. However Banksia Woodlands are generally considered to have been excluded from burning, except where routes between significant areas passed through the woodlands.
Certain fire regimes are inappropriate for the long-term survival of the ecological community and these are a major threat to the diversity, viability and long-term conservation of communities, habitats and populations of many species on the Swan Coastal Plain. These fire effects are the result of cool-season prescribed burning, low intensity and high frequency of fires. While many plant taxa and ecosystems are resilient to a range of fire regimes, Banksia Woodlands and some component species have specific fire regime requirements including fire-free intervals sufficient to allow a build up of seed resources for species susceptible to fire, and sensitivity of geophytic life forms to cool season fires. It is unlikely that any single fire prescription is optimal for all species (Burrows, 2008, Burrows et al., 2008).
More recently, fires have occurred as a result of fire management practices, escapes from prescribed burning operations, arson, and accidental ignition from a range of sources. As a result there has been a fire regime change, with a skewed distribution of fire frequencies (<7 yr fire frequencies are overrepresented).
Higher frequency fire regimes and fire management practices that result in burning during the growing season (late autumn to late spring) and during the seeding season (for most native species in Banksia Woodlands this is from November to December) result in the following changes to Banksia Woodlands (Fisher et al., 2009 a, b; Stevens et al., 2016):
Structural change, e.g. reduction in canopy cover
Increase in weed abundance, diversity and a decrease in native plant diversity and density
Feedback loops, promoting weed species e.g. perennial veldtgrass Ehrharta calycina which is highly flammable and promotes further fires.
The richness and diversity of fauna taxa is generally maximised by avoiding widespread intense bushfires and by maintaining a diversity of post-fire vegetation successional stages to provide habitat diversity (Bamford and Roberts, 2003). The fire responses of native fauna will also vary depending on the extent of, and interaction of fire with, habitat fragmentation and other ecological disturbances (for example the effects of weeds, disease and introduced animals). The response of reptiles to fire in the region has been found to be dependent on vegetation type and fire ages with some species disadvantaged by current prescribed burning practices (Valentine et al., 2012; DPaW, 2014).
Areas of remnant Banksia Woodlands that are small in scale and isolated from other remnants are also particularly sensitive to fire. A high intensity fire that affects the entirety of such a remnant may result in changes in structure of the ecological community, and/or the loss of populations of rare and endemic flora, due to depressed seeding rates or impacts of weeds. Such remnants also tend to experience higher impediments to post-fire recovery, such as kangaroo grazing and invasion of weeds (Fisher et al., 2009a, b; DPaW, 2014).
Plant pathogens (causing dieback)
‘Dieback’ here generally refers to the effects of a plant disease caused by the water mould Phytophthora cinnamomi and other Phytophthora species, although it can be related to a number of plant pathogens. Other common pathogens affecting the Banksia Woodlands ecological community include aerial cankers (e.g. Botryosphaeria ribis), gall rust (Uromycladium tepperianum) (restricted to only some Acacia species) and the native parasitic honey fungus (Armillaria luteobubalina).
The consequences of infection range from localised infection affecting one or more individual plants, to a dramatic modification of the structure and composition of the native plant communities; a significant reduction in primary productivity; and, for dependent flora and fauna, habitat loss and degradation. For Banksia Woodlands, impacts are typically towards the severe extreme of this range.
Dieback disease caused by Phytophthora cinnamomi continues to spread and affect the distribution and abundance of many native southwest Australian plant species and their associated fauna. This plant pathogen and a number of related Phytophthora species present a significant threat to the health and vitality of many ecosystems on the Swan Coastal Plain, including the Banksia Woodlands. Phytophthora cinnamomi can alter species composition and ecosystem functioning, by impacting susceptible species and vegetation types, some of which may be rare or threatened, and by increasing the vulnerability of impacted areas to invasion by weeds (DPaW, 2014) through opening up of the canopy and creation of soil voids.
Transmission of plant pathogens occurs through various vectors such as humans and kangaroos, and on larger scales, through contaminated vehicles and machinery. Effective hygiene practices can help to manage human and mechanical transmission.
Most exotic plant species of the Banksia Woodlands are herbs and grasses and originate from the Mediterranean Basin, California and South Africa (Dodd and Griffin, 1989; Gibson et al., 1994; Stevens et al., 2016). There are many herb and grass weeds in Banksia Woodlands with this system being vulnerable to new weeds due to their proximity to major population centres (Stevens et al., 2016).
The most common and widespread weeds include Gladiolus caryophyllaceus, Freesia refracta, Ursinia anthemoides, Hypochaeris glabra and Briza maxima, and perennial veldtgrass (Ehrharta calycina). Gladiolus caryophyllaceus almost exclusively occurs in Banksia Woodlands.
The weed species with the greatest effect on community composition are freesias, perennial veldtgrass and Gladiolus caryophyllaceus, as they not only transform the ecological character of the community but they also reduce the diversity of the native shrubs, herbs, sedges and grasses.
564 introduced plants are recorded from the Gnangara groundwater system area, which makes up nearly 30% of all plant taxa in the area (Reaveley et al., 2009). Thirty of these introduced plants are identified and have significant ecological impacts due to real or potential invasiveness.
In areas of significant disturbance, Banksia woodlands in the Perth area are altered structurally by the presence of a perennial grass layer dominated by Ehrharta calycina (perennial veldgrass). However, perennial veldgrass is also present in a significant number of the most intact areas as it was recorded in 23% of the sample points that were located in the most intact areas of each plant community sampled. This grass not only competes with native taxa, but it changes the fuel loads in bushland, resulting in bushland being more prone to arson and promoting higher fire frequencies (Stevens et al., 2016).
Common invasive fauna include the European rabbit (Oryctolagus cuniculus), red fox (Vulpes vulpes), feral cat (Felis catus), black rat (Rattus rattus), house mouse (Mus musculus), rainbow lorikeets (Trichiglossus haematodus), laughing kookaburra (Dacelo novaeguineae) and the introduced honey bee (Apis mellifera) (Reaveley et al., 2009; Stevens et al., 2016). Introduced fauna species affect biodiversity values through habitat modification, predation, grazing and competition.
Whilst native herbivores suppress non-native herbaceous species abundance in Banksia woodlands, non-native herbivores such as the European rabbit promote non-native herbaceous species abundance (Ramalho et al., 2014) as a result of their digging activities that promote germination of the weed soil seed bank (Fisher et al., 2009b; Hopper, 2009; Ramalho et al., 2014).
Mining, exploration and extraction
The extraction of raw materials can result in the loss of vegetation and the introduction and spread of dieback and weeds. Demand for basic raw materials such as gravel, shale, clay, sand, limestone and rock for construction and infrastructure development will increase in the future to support population growth (DPaW, 2014; EPA, 2015). Extraction of mineral sands, in particular, can result in the removal of and/or disturbance to Banksia Woodlands, due to their association with the sand dune systems.
Other disturbances to patches
Other disturbances to patches are particularly common in the urban and peri-urban context (Stenhouse, 2004; Ramalho, 2014). Common anthropogenic disturbances to urban remnants include the influx of exotic plant species, especially in the understorey vegetation, dumped rubbish, access by unauthorised vehicles, paths from trampling through the vegetation, illegal cutting of vegetation, firewood collections, bare patches of ground where vegetation cover has been destroyed, erosion, feral animals and domestic animals (Stenhouse, 2004; DPaW, 2014). These impacts are likely to spread out into remnants that are currently in peri-urban and rural areas, with future urban development plans for the Swan Coastal Plain.
Key threatening processes
EPBC-listed key threatening processes relevant to this ecological community as at March 2016 are:
Competition and land degradation by rabbits
Dieback caused by the root-rot fungus (Phytophthora cinnamomi)