Using ants to monitor changes within and surrounding the endangered Monsoon Vine Thickets of the tropical Dampier Peninsula, north Western Australia
Introduction
Tropical Western Australia sits within the monsoonal zone and experiences a climate with high temperatures, a dry season and rainfall that is concentrated in the summer wet season, when approximately 90% of the annual rainfall is received. The vegetation of this region of Australia is predominantly comprised of savanna termed Pindan woodland, with a sparse canopy of trees overlying a lower layer of grasses. Embedded in this woodland are numerous small and isolated semi-deciduous monsoon vine thicket (tropical forest) patches (MVT) (Bowman et al., 2010), where trees form a closed canopy which contrasts with the open canopy of the surrounding Pindan woodland. The position of these rainforests is determined by the presence of permanent water or a topography that provides shelter from desiccation and wildfire (Bowman et al., 1991, McKenzie, 1991, Russell-Smith, 1991).
The MVT patches found in the Kimberley of north-western Australia are floristically and structurally less diverse than the humid rainforests of Queensland, 2000 km to the east, but still have a closed canopy of broad-leaved tree and vine species and a high tree richness, with many trees having buttressed roots and a complex structure (Andersen and Burbidge, 1991, Andersen and Majer, 1991, Andersen, 1992, Andersen and Reichel, 1994, Black et al., 2010). Boundaries between the rainforest and savanna in tropical north-western Australia are quite abrupt (Black et al., 2010, Fisher and Beames, 2013).
Seventy-nine discrete and naturally fragmented MVT patches occur along the Dampier Peninsula of the north-west Kimberley region of Western Australia (Fig. 1), and are found in swales and on the leeward side of the often extensive Holocene dune systems. These naturally fragmented and restricted MVT patches are considered a ‘Threatened Ecological Community’ (TEC) and ‘Endangered’ under the Commonwealth (EPBC Act, 1999) and as ‘Vulnerable’ by the State of Western Australia. The MVT patches are culturally significant and are the primary source of indigenous bush foods, medicines, tools, ceremonial areas and law grounds of indigenous people. Such MVT patches cover less than 0.001% of the Dampier Peninsula (Bellfield et al., 2012), and have a scattered distribution there from Broome north to One Arm Point and then on the western side of King Sound as far south as Goodenough Bay. Plant composition in these MVT patches varies, and includes semi-deciduous plants. Ninety-four per cent of the patches are smaller than 100 ha, and the mode patch size is 10 ha, with the median patch size being 32 ha (Black et al., 2010). These small aggregations of rainforest function as botanical refugia and contain nearly a quarter of all known Dampier Peninsula plant species, with some plant taxa being restricted to MVT patches or even to an individual patch (Black et al., 2010).
MVT patches are prone to high impact from disturbances such as fire, aridification, temperature variations, and introduced species, particularly at the edges, all of which can result in changes to micro-climates and reduced availability of energy, water and nutrients. This in turn leads to changes in the interactions between species found in these environments, and many plants and animals can be affected (Brodie et al., 2012, Pellissier et al., 2013, Struebig et al., 2011).
A popular method of measuring processes of change in Australia is the analysis of ant richness, diversity, abundance and community composition, as ants have been found to be good indicators of ecosystem change (Hoffmann and Andersen, 2003, Andersen and Majer, 2004). Hoffmann and Andersen (2003) analysed 45 studies utilising quantitative data measuring responses of ant communities to disturbance and found virtually all studies used pitfall traps and counts of ant species as abundance data, utilising transformed data to reduce distortions caused by large numbers of ants falling into traps placed near colony entrances or along foraging trails (Andersen, 1990). Not only are ants good bioindicators due to their species composition reflecting deviations in the biological integrity of ecosystems, they can also be used to track the efficacy of remediation efforts and management of the environment (Andersen and Majer, 2004). Since ants are particularly sensitive to habitat structural change and alteration in land management, this will produce a measurable effect on ant communities, an analysis of which can provide different and complementary data to that obtained from the measurement of vegetation parameters alone (Andersen and Majer, 2004).
In the Dampier Peninsula, both patch type and fire may influence how successfully MVT patches can compete with the surrounding savanna, and, as outlined above, the analysis of ant community response to these two factors is a useful tool in evaluating this success. Andersen and Burbidge (1991) have characterised the ant fauna of one MVT patch in Western Australia. The impact of fire on the ant fauna of monsoonal savanna and rainforest edges has been reviewed by Andersen et al. (2012) and the impact of invasive ants, most notably Pheidole megacephala, has been investigated in rainforests of the Northern Territory (Hoffmann et al., 1999). Hoffmann and Saul (2010) have also reviewed the impact of the introduced ant, Anoplolepis gracilipes, in northeast Arnhem Land. This paper provides further information on ant communities and the impact of fire and invasive ants in the Western Australian MVT patches.
In addition to the above, we outline some lessons concerning the involvement of local indigenous people in the monitoring and management of such patches. Traditional Owners recognise that ants have a role in cleaning up the fruits and seeds and maintaining the health of MVT patches. Traditional Owners and Indigenous Ranger Groups from Bardi Jawi and Nyul Nyul Ranger Groups utilised their cultural knowledge in field site selection both in the MVT patches studied and in the demarcation of the inside and edge of the MVT patch and choice of Pindan woodland sites. Site selection reflected areas where Management Plans have been developed and endorsed by Traditional Owners (Environs Kimberley and Bardi Jawi Rangers, 2012). Thus, the results provide data which can be used in future studies where the implementation of management actions have provided an opportunity to monitor environmental changes, using protocols developed with this aim in mind (Fisher and Beames, 2013).
This paper has four aims:
- (i)
To assess the ant richness, diversity, abundance and community composition of five MVT patches of the Dampier Peninsula by sampling sites inside (I) and at the edge (E) of the MVT patch, and outside (O) in the surrounding Pindan woodland.
- (ii)
To determine if there is a relationship between ant community composition, environmental variables and fire history.
- (iii)
To investigate whether invasive ant species are present and, if present, whether these have had an impact on the ant community composition.
- (iv)
To work together with Indigenous rangers and Traditional Owners to undertake monitoring of MVT patches and provide details about the benefits of this initiative.
Section snippets
Study area
The study was conducted in five discrete MVT patches (11 ha–56 ha) situated north of the 750 mm isohyet (Table 1). Traditional knowledge was utilised in patch site selection and in the demarcation between the three habitats per patch, namely inside (I) and edge (E) of the MVT patches, and the surrounding (O) Pindan savanna woodland. Although edges can be quite diffuse in some areas, the ones chosen had well defined edges. The MVT patches chosen are managed using traditional and modern ecological
Alpha diversity
In total 7342 ants were collected across the 15 study sites, over the two sampling seasons. These belonged to seven subfamilies: Myrmicinae (31 spp.), Dolichoderinae (13 spp.), Formicinae (21 spp.), Ponerinae (7 spp.), Ectatomminae (6 spp.), Cerapachyinae (2 spp.) and Pseudomyrmecinae (1 spp.). A total of 76 species were recorded from pitfall traps with 5 additional species recorded by hand collection, making 81 species in total (Table 2). Thirty-four species were found in one habitat type only
Discussion
The overall results from the sampling reveal a mainly generalised ant fauna that is largely typical of the northern Pilbara and the Torresian zone. Conspicuously absent are the small, specialised myrmicines (e.g., Dacetini (Bolton, 2000) or Basicerotini) and ponerines characteristic of the lush Queensland rainforest. Characteristically, Queensland humid rainforests have gradients of regrowth between rainforest and savanna (Van Ingen et al., 2008), and these may be significant in this context.
Conclusion
Our results have underlined the importance of understanding interacting processes and drivers of change, including fire history, vegetation structure and environmental variables at the regional scale, both within the three habitats (I, E, O) and between the five MVT patches, in determining both ant species richness and abundance and the value of this knowledge to enhance management practices. Our findings are also consistent with other studies (Kaspari et al., 2003, Parr et al., 2005, Andersen,
Acknowledgements
The Environs Kimberley West Kimberley Nature Project, run by ecologist Louise Beames, has received funding from the State NRM Western Australia, and is funded by Rangelands NRM WA through Caring for Our Country (Australian Federal Government). Bardi Jawi and Nyul Nyul Rangers monitor manage and protect their natural, cultural land and sea resources. They are supported by Traditional Owners, funded through Working on Country, and facilitated by the Kimberley Land Council. Dr Judith Fisher
References (71)
Evolution of co-management: role of knowledge generation, bridging organisations and social learning
J. Environ. Manage.
(2009)- et al.
Climate change and tropical biodiversity: a new focus
Trends Ecol. Evol.
(2012) - et al.
Altered vegetation structure and composition linked to fire frequency and plant invasion in a biodiverse woodland
Biol. Conserv.
(2009) - et al.
How large is large enough for insects? Forest fragmentation effects at three scales
Acta Oecol.
(2005) - et al.
Ants, Standard Methods for Measuring and Monitoring Biodiversity
(2000) Immediate and longer-term effects of fire on seed predation by ants in sclerophyllous vegetation in south-eastern Australia
Aust. J. Ecol.
(1988)The use of ant communities to evaluate change in Australian terrestrial ecosystems: a review and a recipe
Proc. Ecol. Soc. Aust.
(1990)The rainforest ant fauna of the northern Kimberley region of Western Australia (Hymenoptera: Formicidae)
J. Aust. Entomol. Soc.
(1992)Not enough niches: non equilibrium processes promoting species co-existence in ant communities
Austral Ecol.
(2008)- et al.
The ants of a vine thicket near Broome: a comparison with the northwest Kimberley
J. Roy. Soc. Western Aust.
(1991)